Diagnosis Registry Research Articles

Differential Ultrasound Rates Mirror Sex Disparities in Thyroid Cancer.

Background: Expanding ultrasound use has increased the detection of thyroid cancer. Incidence has always been higher among females, a disparity that has grown over time. The sex difference in thyroid cancer is understudied in the context of diagnostic testing, particularly among privately insured adults in whom thyroid cancer is most common. We evaluated the association between thyroid ultrasound, fine needle aspiration biopsy (FNAB), and cancer incidence by sex in a large, integrated health system. Methods: This longitudinal retrospective cohort study included Kaiser Permanente of Washington enrollees aged 18 and over who underwent thyroid ultrasound from 1997 to 2019. Data included electronic billing claims for patients linked to tumor registry diagnoses. We estimated (1) annual overall ultrasound, FNAB, and cancer incidence rates; (2) the proportion of ultrasound requiring FNAB; and (3) cancer diagnoses per FNAB. A Poisson model with offset determined the relationship between sex and the proportion of ultrasound requiring FNAB adjusting for patient and sociodemographic characteristics. Results: A total of 33,589 patients underwent ultrasound (78% females; mean age 56). Ultrasound rates per 100,000 covered lives, defined as insured individuals per year, increased five-fold among males (111.11-490.97) and >four-fold among females (382.27-1331.14) between 1997 and 2019. FNAB rates also increased over time (rates per 100,000: 174.09-430.37 in females vs. 58.38-189.13 in men). Overall, FNAB rates per ultrasound changed little over time, and FNAB per ultrasound was greater in males compared with females (Adj rate ratio = 1.06 [confidence interval 1.01-1.11]). Cancer incidence was higher in females over the study period, but cancer incidence per FNAB was similar between sexes (both 0.06, p = 0.4). Conclusions: Sex disparities in thyroid ultrasound rates are stark and are a likely driver of sex disparities in thyroid cancer incidence. Interestingly, ultrasound-triggered FNAB was more common in males and changed little over time, challenging the prevailing understanding that females have much higher rates of thyroid cancer. Although the population-based differences between sexes for FNAB and cancer were large, the differences among people who had ultrasound were small.

Use of specialist health care among Norwegian-born children with one immigrant parent

Abstract Background An increasing number of children grow up with one immigrant and one native born parent. We know little about health in this group. Our aim was to compare the risk of receiving diagnoses of physical health conditions in specialist health care among Norwegian-born children with one immigrant parent compared to Norwegian background children. Methods Data from Medical Birth Registry of Norway were linked to data from the Norwegian Patient Registry (diagnoses in specialist health care), and Statistics Norway (parental country of origin, education, and household income). We estimated hazard ratios (HR) with Cox proportional hazard regressions, with Norwegian background children as reference category, adjusted for sex, birth year, parental education, and household income. Results Children with an immigrant parent had lower hazards than Norwegian background children of diagnoses of most types of infections, but higher hazards of some, including tuberculosis, and skin infections. They also had higher hazards than Norwegian background children of receiving diagnoses of obesity and other nutritional deficiencies, and skin diseases, but lower hazards for other conditions, including chronic lower respiratory disease, disease of the digestive system, and visual or hearing impairment. The hazards of diagnoses of any neurological condition, sleeping disorders and epilepsy were lower among children with an immigrant mother, but higher among those with an immigrant father. Differences between children with one immigrant parent and Norwegian background could not be explained by differences in parental education or household income. Conclusions Children of one immigrant parent do not overall use specialist health care for physical health issues more than other children. Key messages • Awareness of diagnoses given most frequently to children of one immigrant parent can help promote timely diagnoses and adequate use of services, which is important for prevention of later illness. • Barriers to health care use among children of immigrants should be further researched.

A phenome-wide association study of polygenic scores for selected childhood cancer: Results from the UK Biobank

The aim of this study was to scan phenotypes in adulthood associated with polygenic risk scores (PRS) for childhood cancers with well-articulated genetic architectures: acute lymphoblastic leukemia (ALL), Ewing sarcoma, and neuroblastoma to examine genetic pleiotropy. Furthermore, we aimed to determine which single nucleotide polymorphisms (SNPs) could drive associations. Per-SNP summary statistics were extracted for PRS calculation. Participants with white British ancestry were exclusively included for analyses. SNPs were queried from the UK biobank genotype imputation data. Records from the cancer registry, death registry, and inpatient diagnoses were abstracted for phenome-wide scans. Firth logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) alongside corresponding p-values, adjusting for age at recruitment and sex. A total of 244,332 unrelated white British participants were included. We observed a significant association between ALL-PRS and ALL (OR: 1.20e+24, 95% CI: 9.08e+14-1.60e+33). In addition, we observed a significant association between high-risk neuroblastoma PRS and nonrheumatic aortic valve disorders (OR: 43.9, 95% CI: 7.42-260). There were no significant phenotype associations with Ewing sarcoma and neuroblastoma PRS. Regarding individual SNPs, rs17607816 increased the risk of ALL (OR: 6.40, 95% CI: 3.26-12.57). For high-risk neuroblastoma, rs80059929 elevated the risk of atrioventricular block (OR: 3.04, 95% CI: 1.85-4.99). Our findings suggest that individuals with genetic susceptibility to ALL may face a lifelong risk for developing ALL, along with a genetic pleiotropic association between high-risk neuroblastoma and circulatory diseases.

Prevalence and Determinants of Diagnosed Dementia: A Registry Linkage Study Linking Diagnosis of Dementia in the Population-Based HUNT Study to Registry Diagnosis of Dementia in Primary Care and Hospitals in Norway.

A timely diagnosis of dementia can be beneficial for providing good support, treatment, and care, but the diagnostic rate remains unknown and is probably low. To determine the dementia diagnostic rate and to describe factors associated with diagnosed dementia. This registry linkage study linked information on research-based study diagnoses of all-cause dementia and subtypes of dementias, Alzheimer's disease, and related dementias, in 1,525 participants from a cross-sectional population-based study (HUNT4 70+) to dementia registry diagnoses in both primary-care and hospital registries. Factors associated with dementia were analyzed with multiple logistic regression. Among those with research-based dementia study diagnoses in HUNT4 70+, 35.6% had a dementia registry diagnosis in the health registries. The diagnostic rate in registry diagnoses was 19.8% among home-dwellers and 66.0% among nursing home residents. Of those with a study diagnosis of Alzheimer's disease, 35.8% (95% confidence interval (CI) 32.6-39.0) had a registry diagnosis; for those with a study diagnosis of vascular dementia, the rate was 25.8% (95% CI 19.2-33.3) and for Lewy body dementias and frontotemporal dementia, the diagnosis rate was 63.0% (95% CI 48.7-75.7) and 60.0% (95% CI 43.3-75.1), respectively. Factors associated with having a registry diagnosis included dementia in the family, not being in the youngest or oldest age group, higher education, more severe cognitive decline, and greater need for help with activities of daily living. Undiagnosed dementia is common, as only one-third of those with dementia are diagnosed. Diagnoses appear to be made at a late stage of dementia.

Applications of Natural Language Processing for Automated Clinical Data Analysis in Orthopaedics.

Natural language processing is an exciting and emerging field in health care that can transform the field of orthopaedics. It can aid in the process of automated clinical data analysis, changing the way we extract data for various purposes including research and registry formation, diagnosis, and medical billing. This scoping review will look at the various applications of NLP in orthopaedics. Specific examples of NLP applications include identification of essential data elements from surgical and imaging reports, patient feedback analysis, and use of AI conversational agents for patient engagement. We will demonstrate how NLP has proven itself to be a powerful and valuable tool. Despite these potential advantages, there are drawbacks we must consider. Concerns with data quality, bias, privacy, and accessibility may stand as barriers in the way of widespread implementation of NLP technology. As natural language processing technology continues to develop, it has the potential to revolutionize orthopaedic research and clinical practices and enhance patient outcomes.

Higher incidence of paediatric inflammatory bowel disease by increasing latitude in Norway, but stable incidence by age.

To examine possible geographical and temporal differences in the incidence of childhood-onset inflammatory bowel disease (IBD) in Norway, motivated by previous research indicating relevant environmental factors explaining changing epidemiology. We analysed data from children born in Norway from 2004 to 2012 (n = 541 036) in a registry-based nationwide study. After validating registry diagnoses against medical records, we defined IBD as ≥2 entries of International Classification of Diseases, 10th revision (ICD-10) codes K50, K51 and K52.3 in the Norwegian Patient registry. We estimated hazard ratios (HR) for IBD across four geographical regions with a south-to-north gradient and the incidence by period of birth. By the end of follow-up on 31 December 2020, 799 IBD diagnoses were identified (Crohn's disease: n = 465; ulcerative colitis, n = 293, IBD: unclassified, n = 41). Compared to children in the southernmost region, there was almost a two-fold HR for IBD in children in the most Northern region (HR = 1.94, 95% Cl = 1.47-2.57; Mid region: HR = 1.68, 95% CI = 1.29-2.19, ptrend <0.001). These estimates remained largely unchanged after adjustment for potential confounding factors. The cohorts born in 2004-2006 and 2010-2012 had comparable cumulative incidences, with a slightly higher incidence for those born in 2007-2009. We observed an increase in the risk of IBD by increasing latitude which may suggest that environmental factors influence the development of IBD, although non-causal explanations cannot be ruled out.

A practical guide to the development and deployment of deep learning models for the orthopaedic surgeon: Part III, focus on registry creation, diagnosis,and data privacy.

Deep learning is a subset of artificial intelligence (AI) with enormous potential to transform orthopaedic surgery. As has already become evident with the deployment of Large Language Models (LLMs) like ChatGPT (OpenAI Inc.), deep learning can rapidly enter clinical and surgical practices. As such, it is imperative that orthopaedic surgeons acquire a deeper understanding of the technical terminology, capabilitiesand limitations associated with deep learning models. The focus of this series thus far has been providing surgeons with an overview of the steps needed to implement a deep learning-based pipeline, emphasizing some of the important technical details for surgeons to understand as they encounter, evaluateor lead deep learning projects. However, this series would be remiss without providing practical examples of how deep learning models have begun to be deployed and highlighting the areas where the authors feel deep learning may have the most profound potential. While computer vision applications of deep learning were the focus of Parts I and II, due to the enormous impact that natural language processing (NLP) has had in recent months, NLP-based deep learning models are also discussed in this final part of the series. In this review, three applications that the authors believe can be impacted the most by deep learning but with which many surgeons may not be familiar are discussed: (1) registry construction, (2) diagnostic AIand (3) data privacy. Deep learning-based registry construction will be essential for the development of more impactful clinical applications, with diagnostic AI being one of those applications likely to augment clinical decision-making in the near future. As the applications of deep learning continue to grow, the protection of patient information will become increasingly essential; as such, applications of deep learning to enhance data privacy are likely to become more important than ever before. Level of Evidence: Level IV.

Validation of the Finnish Care register for Health Care diagnoses for preeclampsia, gestational diabetes and preterm delivery

BackgroundCentrally collected Finnish national health register data on adverse pregnancy outcomes are available for research, but the validity of the data is largely unknown. Our aim was to compare the diagnoses of preeclampsia (PE), gestational diabetes (GDM), and preterm delivery from hospital records with the registry based diagnoses from the Finnish Care Register for Health Care (FCR). Data on gestational age at delivery from the Medical Birth Registry (MBR) was also studied. MethodsThe Finnish Genetics of Pre-eclampsia Consortium (FINNPEC) Study cohort was used as a data source. Each diagnosis was ascertained from electronic hospital records. The validity of diagnoses obtained by record linkage of FCR and MBR was assessed against the classification previously confirmed independently by a research nurse and a study physician. ResultsSensitivity of PE diagnoses in FCR was 80.3 % (95 % CI 78.3 % to 82.2 %) andspecificity 95.3 % (95 % CI 93.9 % to 96.4 %). Sensitivity for GDM was 64.1 % (95 % CI: 58.7 % − 69.3 %) and specificity 98.5 % (95 % CI: 97.9 % − 98.9 %), whereas sensitivity and specificity for preterm delivery were 32.4 % (95 % CI: 29.0 % − 36.0 %) and 99.7 % (95 % CI: 99.3 % − 99.9 %). Sensitivity of preterm delivery in the MBR was 99.1 % and specificity 99.9 %. ConclusionsFCR registry diagnoses for PE have satisfactory sensitivity and high specificity. Diagnoses for GDM and preterm delivery have lower sensitivity limiting their use in studies, and data from MBR should be preferred when studying preterm deliveries.

Pancreatic cancer symptom trajectories from Danish registry data and free text in electronic health records.

Pancreatic cancer is one of the deadliest cancer types with poor treatment options. Better detection of early symptoms and relevant disease correlations could improve pancreatic cancer prognosis. In this retrospective study, we used symptom and disease codes (ICD-10) from the Danish National Patient Registry (NPR) encompassing 6.9 million patients from 1994 to 2018,, of whom 23,592 were diagnosed with pancreatic cancer. The Danish cancer registry included 18,523 of these patients. To complement and compare the registry diagnosis codes with deeper clinical data, we used a text mining approach to extract symptoms from free text clinical notes in electronic health records (3078 pancreatic cancer patients and 30,780 controls). We used both data sources to generate and compare symptom disease trajectories to uncover temporal patterns of symptoms prior to pancreatic cancer diagnosis for the same patients. We show that the text mining of the clinical notes was able to complement the registry-based symptoms by capturing more symptoms prior to pancreatic cancer diagnosis. For example, 'Blood pressure reading without diagnosis', 'Abnormalities of heartbeat', and 'Intestinal obstruction' were not found for the registry-based analysis. Chaining symptoms together in trajectories identified two groups of patients with lower median survival (<90 days) following the trajectories 'Cough→Jaundice→Intestinal obstruction' and 'Pain→Jaundice→Abnormal results of function studies'. These results provide a comprehensive comparison of the two types of pancreatic cancer symptom trajectories, which in combination can leverage the full potential of the health data and ultimately provide a fuller picture for detection of early risk factors for pancreatic cancer.

Determining the sensitivity of emergency dispatcher and paramedic diagnosis of stroke: statewide registry linkage study.

IntroductionCorrectly identifying people with suspected stroke is essential for ensuring rapid treatment. Our aims were to determine the sensitivity of emergency dispatcher and paramedic identification of patients with stroke, the factors associated with correct identification, and whether there were any implications for hospital arrival times.MethodsObservational study using patient‐level data from the Australian Stroke Clinical Registry (2015–2017) linked with ambulance and emergency department records for the state of Victoria. The registry diagnosis was the reference standard to compare with the provisional diagnoses made by emergency services personnel classified as “suspected” and “not suspected” stroke/transient ischemic attack (TIA). Multivariable logistic and quintile regressions were used to determine factors associated with correct identification and timely arrival to hospital.ResultsOverall, 4717 (64%) were matched to ambulance transport records (median age: 73 years, 43% female). Stroke/TIA was suspected in 56% of registrants by call‐takers and 69% by paramedics. Older patients (75+ years) (adjusted odds ratio [aOR]: 0.61; 95% confidence interval [CI]: 0.49–0.75), females (aOR: 0.86; 95% CI: 0.75–0.99), those with severe stroke or intracerebral hemorrhage were less often suspected as stroke. Cases identified as stroke had a shorter arrival time to hospital (unadjusted median minutes: stroke, 54 [43, 72] vs not stroke, 66 [51, 89]).ConclusionsEmergency dispatchers and paramedics identified over half of patients with stroke in the prehospital setting. Important patient characteristics, such as being female and those having a severe stroke, were found that may enable refinement of prehospital ambulance protocols and dispatcher/paramedic education. Those correctly identified as stroke, arrived earlier to hospital optimizing their chances of receiving time‐critical treatments.

Abstract 242: Assessing Variation In Data Captured By Clinical Registries And The Electronic Health Record: Clinical Phenotypes Of Patients Undergoing Cardiac Catheterization

Background: Cardiovascular registries are built through expert abstraction of summary clinical information often represented as binary flags used to define diagnosis status. We considered underlying variation in clinical data obtained from the electronic health record (EHR) with respect to binary flags denoting presence or absence of cardiometabolic disease in a clinical data registry. Methods: We linked data abstracted for percutaneous coronary intervention (PCI) visits registered in NCDR CathPCI® with EHR data from Yale New Haven Hospital. We identified registry flags indicating diagnoses of hyperlipidemia (HLD), hypertension (HTN) and diabetes mellitus (DM) and evaluated EHR data with respect to registry diagnosis status. Results: Among 3,473 patients, mean age was 67 ± 12 years and 960 (28%) patients were female. HLD, HTN and DM registry diagnoses were present (+) in 2694 (78%), 2833 (82%) and 1358 (39%) patients, and absent (-) in 779 (22%), 640 (18%) and 2115 (61%) patients. EHR data revealed mean baseline total cholesterol of 168 ± 51 mg/dL (+HLD) vs 182 ± 44 mg/dL (-HLD), presenting systolic blood pressure of 130 ± 19 mmHg (+HTN) vs 120 ± 17 mmHg (-HTN), and A1c of 7.9% ± 1.8% (+DM) vs 5.8% ± 0.7% (-DM). Prescription histories of patients with HLD, HTN or DM diagnoses included 1 or more relevant drug class in 2409 (89%), 2620 (92%) and 1222 (90%) patients, and 2 or more drug classes in 450 (17%), 2020 (71%) and 820 (60%) patients. Patients without these diagnoses had been prescribed 1 or more drug class in 268 (34%), 223 (35%) and 55 (3%) patients, and 2 or more drug classes in 28 (4%), 71 (11%) and 9 (0.4%) patients (Figure 1). Conclusions: Binary diagnosis flags used in cardiovascular registries do not capture the full variation of clinical phenotypes underlying cardiometabolic disease processes. Harnessing comprehensive EHR data may provide a fuller picture of cardiometabolic disease, enhance quality and utility of national registries and improve outcomes research.

Developing an algorithm across integrated healthcare systems to identify a history of cancer using electronic medical records.

Tumor registries in integrated healthcare systems (IHCS) have high precision for identifying incident cancer but often miss recently diagnosed cancers or those diagnosed outside of the IHCS. We developed an algorithm using the electronic medical record (EMR) to identify people with a history of cancer not captured in the tumor registry to identify adults, aged 40-65 years, with no history of cancer. The algorithm was developed at Kaiser Permanente Colorado, and then applied to 7 other IHCS. We included tumor registry data, diagnosis and procedure codes, chemotherapy files, oncology encounters, and revenue data to develop the algorithm. Each IHCS adapted the algorithm to their EMR data and calculated sensitivity and specificity to evaluate the algorithm's performance after iterative chart review. We included data from over 1.26 million eligible people across 8 IHCS; 55 601 (4.4%) were in a tumor registry, and 44848 (3.5%) had a reported cancer not captured in a registry. The common attributes of the final algorithm at each site were diagnosis and procedure codes. The sensitivity of the algorithm at each IHCS was 90.65%-100%, and the specificity was 87.91%-100%. Relying only on tumor registry data would miss nearly half of the identified cancers. Our algorithm was robust and required only minor modifications to adapt to other EMR systems. This algorithm can identify cancer cases regardless of when the diagnosis occurred and may be useful for a variety of research applications or quality improvement projects around cancer care.

Positive predictive value and sensitivity of ICD-9-CM codes for identifying pediatric leukemia.

To facilitate community-based epidemiologic studies of pediatric leukemia, we validated use of ICD-9-CM diagnosis codes to identify pediatric leukemia cases in electronic medical records of six U.S. integrated health plans from 1996-2015 and evaluated the additional contributions of procedure codes for diagnosis/treatment. Subjects (N=408) were children and adolescents born in the health systems and enrolled for at least 120 days after the date of the first leukemia ICD-9-CM code or tumor registry diagnosis. The gold standard was the health system tumor registry and/or medical record review. We calculated positive predictive value (PPV) and sensitivity by number of ICD-9-CM codes received in the 120-day period following and including the first code. We evaluated whether adding chemotherapy and/or bone marrow biopsy/aspiration procedure codes improved PPV and/or sensitivity. Requiring receipt of one or more codes resulted in 99% sensitivity (95% confidence interval [CI]: 98-100%) but poor PPV (70%; 95% CI: 66-75%). Receipt of two or more codes improved PPV to 90% (95% CI: 86-93%) with 96% sensitivity (95% CI: 93-98%). Requiring at least four codes maximized PPV (95%; 95% CI: 92-98%) without sacrificing sensitivity (93%; 95% CI: 89-95%). Across health plans, PPV for four codes ranged from 84-100% and sensitivity ranged from 83-95%. Including at least one code for a bone marrow procedure or chemotherapy treatment had minimal impact on PPV or sensitivity. The use of diagnosis codes from the electronic health record has high PPV and sensitivity for identifying leukemia in children and adolescents if more than one code is required.

Unique challenges and outcomes of young women with breast cancers from a tertiary care cancer centre in India

BackgroundYoung (≤40 years) breast cancers (YBC) are uncommon, inadequately represented in trials and have unique concerns and merit studying. MethodsThe YBC treated with a curative intent between 2015 and 2016 at our institute were analysed. ResultsThere were 1228 patients with a median age of 36 (12–40) years; 38 (3.1%) had Stage I, 455 (37.1%) - II, 692 (56.3%) –III, and remaining 43 (3.5%) Stage IV (oligo-metastatic) disease; 927 (75.5%) were node positive; 422 (34.4%) were Triple negatives (TNBC), 331 (27%) were HER-2 positive. There were 549 (48.2%) breast conservations and 591 (51.8%) mastectomies of which 62 (10.4%) underwent breast reconstruction. 1143 women received chemotherapy, 617 (53.9%) received as neoadjuvant and 142 (23.1%) had pathological complete response; 934 (81.9%) received adjuvant radiotherapy. At the median follow-up of 48 (0–131) months, 5-year overall and disease-free survival was 79.6% (76.8–82.5) and 59.1% (55.8–62.6). For stage I, II, III and IV, the 5-year overall-survival was 100%, 86.7% (82.8–90.6), 77.3% (73.4–81.2), 69.7% (52.5–86.9) and disease-free survival was 94% (85.9–100), 65.9% (60.3–71.5), 55% (50.5–59.5), and 29.6% (14–45.2) respectively. On multivariate analysis, TNBC and HER-2+ subgroups had poorer survival (p = 0.0035). 25 patients had BRCA mutations with a 5-year DFS of 65.1% (95% CI:43.6–86.6). Fertility preservation was administered in 104 (8.5%) patients; seven women conceived and 5 had live births. Significant postmenopausal symptoms were present in 153 (13%) patients. ConclusionMore than half of the YBC in India were diagnosed at an advanced stage with aggressive features leading to suboptimal outcomes. Awareness via national registry and early diagnosis is highly warranted. Menopausal symptoms and fertility issues are prevalent and demand special focus.

Hemoglobin A1c-levels and subsequent risk of depression in individuals with and without diabetes.

It has been suggested that long-term glycemic load as reflected in plasma levels of Glycosylated Hemoglobin, Type A1C (HbA1c) is associated with higher risk of depression, however results have been conflicting. We examined the potential association between HbA1c and risk of depression in a large population-based cohort without baseline diabetes (the Glostrup cohort) defined by either self-reported diabetes, registry diagnosis of diabetes or use of antidiabetic medication at baseline and in a national diabetes cohort (the Danish Adult Diabetes Database). A total of 16,124 middle-aged individuals from the Glostrup cohort and 93,544 patients registered in the Danish Adult Diabetes Database were followed from the first registered HbA1c measurement (1999-2014) for subsequent diagnosis of depression or use of antidepressant medication in nation-wide Danish registers. The association was analyzed using a Cox proportional hazards regression model with HbA1c on both a continuous scale using restricted cubic splines and categorized based on the groups found in the spline model. We adjusted for relevant sociodemographic and clinical variables including previous depression and tested for interaction of both gender, insulin use and diabetes type. During follow-up, 2694 (17%) in the Glostrup cohort and 29,234 (31%) in the diabetes cohort developed depression. In the Glostrup cohort, we found an indication of a positive linear association between HbA1c and depression in women, while no clear association was found in men. In patients with diabetes, we found a U-shaped association between HbA1c and depression in both men and women with the lowest risk estimates for HbA1c levels of 58 mmol/mol (7.5%) in men and of 60 mmol/mol (7.6%) in women. When HbA1c was categorized, men with the highest HbA1c-levels had significantly elevated risk of depression (HRHbA1c>9.4 1.16 (95%CI 1.10-1.23)) after multifactorial adjustment compared to the reference group with HbA1c of 42.1-56.2 mmol/mol (6.0-7.3%). Women in the lowest and highest category of HbA1c had significantly higher risk of depression HRHbA1c<6.0 1.15 (95% CI 1.09-1.22) and HRHbA1c>9.3 1.10 (95% CI 1.04-1.16), respectively, compared to the reference group with HbA1c 42.1-55.0 mmol/mol (7.2-9.3%). There was a significant interaction with gender, but no interaction for insulin use or diabetes type. In a population without baseline diabetes, higher HbA1c levels seemed associated with higher depression risk in women, whereas a U-shaped association was found in patients with known diabetes.

Association of Prenatal Exposure to Valproate and Other Antiepileptic Drugs With Intellectual Disability and Delayed Childhood Milestones

There is concern about neurodevelopmental outcomes associated with prenatal exposure to valproate and other antiepileptic drugs (AEDs) among children of mothers with or without epilepsy. To study the risk of intellectual disability and delayed development in childhood milestones among children of women who used valproate or other AEDs during pregnancy. This population-based cohort study analyzed information on use of AEDs from the Danish National Prescription Registry and register diagnoses from the Danish Psychiatric Central Research Register and Danish National Patient Registry. The study included all live-born singletons in Denmark from January 1, 1997, to December 31, 2011. Data were analyzed in April 2020. Prenatal exposure to maternal valproate and other AEDs. The main measures were adjusted Cox regression estimates of hazard ratios (aHRs) for intellectual disability and a combined outcome of intellectual disability with delayed childhood milestones. A total of 913 302 children (468 708 [51.3%] boys; mean [SD] age, 10.3 [4.4] years and median [interquartile range] age, 10.1 [6.5-14.0] years at final follow-up) were identified and contributed more than 10.2 million person-years of observation, including 580 children exposed to valproate (302 [51.3%] boys). At end of follow-up, 6958 children (0.8%) were identified as having intellectual disability and 14 967 children (1.6%) were identified as having intellectual disability with delayed childhood milestones. Compared with offspring not exposed to valproate prenatally, offspring of women who used valproate during pregnancy had increased risk of intellectual disability (aHR, 4.48; 95% CI, 2.97-6.76) and intellectual disability with delayed childhood milestones (aHR, 6.07; 95% CI, 4.67-7.89). Among mothers with epilepsy, offspring exposed prenatally to valproate had increased risk of intellectual disability (aHR, 1.95; 95% CI, 1.21-3.14) and intellectual disability with delayed childhood milestones (aHR, 3.07; 95% CI, 2.24-4.20) compared with offspring without prenatal exposure. Compared with offspring without prenatal exposure to AEDs, increased risk of intellectual disability was identified in children with prenatal exposure to maternal monotherapy use of carbamazepine (aHR, 3.84; 95% CI, 2.32-6.38), clonazepam (aHR, 2.41; 95% CI, 1.09-5.35), and oxcarbazepine (aHR, 3.70; 95% CI, 2.11-6.51) but not lamotrigine (aHR, 1.33; 95% CI, 0.71-2.48). These findings suggest that prenatal exposure to valproate was associated with increased risk of intellectual disability and delayed childhood milestones. Statistically significant associations were also found for prenatal exposure to other AEDs. These findings suggest that women of childbearing potential may need to be counseled on use of AEDs.

Abstract 5778: Hernia and Ewing sarcoma

Abstract Introduction: Ewing sarcoma (ES) is a rare cancer with an incidence that peaks in adolescence and young adulthood. The tumor occurs in bone and soft tissue and the etiology is unclear. Several studies have observed increased ES risk in children and young adults with hernia, including umbilical and inguinal hernia, suggesting a mechanism for a common causal pathway. A meta-analysis of 3 case-control studies reported a threefold increased risk of ES among hernia patients. The purpose of this study is to examine the association between hernia and ES in a population-based study. Methods: We ascertained Ewing sarcoma cases (age &amp;lt;20 years) from the Danish Cancer Registry (diagnoses 1977-2016; births 1977-2013), then randomly selected controls free of cancer at the date of corresponding cases diagnosis and matched (25:1) by sex and year of birth from the Central Population Register (N=4,160). Ewing sarcoma cases (N=122) were identified with the International Classification of Diseases for Oncology (ICD-O-1 and ICD-O-3 code 9260). Information on birth outcomes and gestational factors were taken from the Medical Birth Registry (1973+) and information on mother's and child's health was taken from the National Patient Register (1977+). We also searched this Register for diagnoses of hernia in the child based on procedure codes (a Nordic coding system from 1977-1995, and an extended version of ICD-10 thereafter). We used conditional logistic regression to determine associations for hernia and ES. Results: 122 ES cases were identified, among whom 6 cases had a diagnosis of hernia (1 of them had hernia diagnosed within the first 3 years of life). Among controls, 203 (6.7%) had a diagnosis of hernia. Of all children diagnosed with hernia, 27 (12.9%) presented hernia during the first year of life and 74 (35.4%) presented during the first 3 years of life. In ES cases there was no increase in risk of hernia (OR= 0.53, 0.17-1.72). Adjustment for presence of other congenital malformations did not change the results for hernia (OR=0.65). Restriction to younger cases (diagnosed ≤15 years) showed similar results for hernia. Conclusions: Previous studies have raised the possibility that inherited genes causing both developmental anomalies and cancer might be involved with the etiology of ES. However, based on registry and hospitalization data, we were not able to confirm previous associations with hernia. Either this study or earlier studies may have been influenced by small sample size. There is also a possibility that previous reports of positive associations are due to recall or publication biases. Citation Format: Di He, Johnni Hansen, Noah Federman, Jorn Olsen, Beate Ritz, Julia E. Heck. Hernia and Ewing sarcoma [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5778.

Adjuvant Systemic Chemotherapy vs Active Surveillance Following Up-front Resection of Isolated Synchronous Colorectal Peritoneal Metastases

To date, there are no data on the value of adjuvant systemic chemotherapy following up-front resection of isolated synchronous colorectal peritoneal metastases. To assess the association between adjuvant systemic chemotherapy and overall survival following up-front resection of isolated synchronous colorectal peritoneal metastases. In this population-based, observational cohort study using nationwide data from the Netherlands Cancer Registry (diagnoses between January 1, 2005, and December 31, 2017; follow-up until January 31, 2019), 393 patients with isolated synchronous colorectal peritoneal metastases who were alive 3 months after up-front complete cytoreductive surgery with hyperthermic intraperitoneal chemotherapy were included. Patients allocated to the adjuvant systemic chemotherapy group were matched (1:1) with those allocated to the active surveillance group by propensity scores based on patient-, tumor-, and treatment-level covariates. Adjuvant systemic chemotherapy, defined as systemic chemotherapy without targeted therapy, starting within 3 months postoperatively. Overall survival was compared between matched groups using Cox proportional hazards regression analysis adjusted for residual imbalance. A landmark analysis was performed by excluding patients who died within 6 months postoperatively. A sensitivity analysis was performed to adjust for unmeasured confounding by major postoperative morbidity. Of 393 patients (mean [SD] age, 61 [10] years; 181 [46%] men), 172 patients (44%) were allocated to the adjuvant systemic chemotherapy group. After propensity score matching of 142 patients in the adjuvant systemic chemotherapy group with 142 patients in the active surveillance group, adjuvant systemic chemotherapy was associated with improved overall survival compared with active surveillance (median, 39.2 [interquartile range, 21.1-111.1] months vs 24.8 [interquartile range, 15.0-58.4] months; adjusted hazard ratio [aHR], 0.66; 95% CI, 0.49-0.88; P = .006), which remained consistent after excluding patients who died within 6 months postoperatively (aHR, 0.68; 95% CI, 0.50-0.93; P = .02) and after adjustment for major postoperative morbidity (aHR, 0.71; 95% CI, 0.53-0.95). Findings of this study suggest that in patients undergoing up-front resection of isolated synchronous colorectal peritoneal metastases, adjuvant systemic chemotherapy appeared to be associated with improved overall survival. Although randomized trials are needed to address the influence of potential residual confounding and allocation bias on this association, results of this study may be used for clinical decision-making in this patient group for whom no data are available.

Abstract C085: Racial/ethnic disparities in pancreatic cancer incidence rates and temporal trends in California (1988-2015)

Abstract Background: Pancreatic cancer is a leading cause of cancer death in the United States. Racial/ethnic disparities in pancreatic cancer incidence exist, with African Americans experiencing the highest incidence. Little is known about incidence and temporal trends in other minorities, especially in Asian Americans, heterogeneous populations that are the fastest-growing minority in US. To better understand pancreatic cancer burden, we examined pancreatic cancer incidence patterns and temporal trends among detailed racial/ethnic populations, including Asian-American subgroups. Methods: We identified a total of 97,665 invasive pancreatic cancer cases (ICD-O-3: C25.0-C25.9) using the California Cancer Registry diagnoses between 1988 and 2015. Cases were grouped into mutually exclusive major racial/ethnic groups of non-Hispanic (NH) white, NH black, Hispanic, NH Asian/Pacific Islander (API), and NH American Indian/Alaska Native (AIAN). Asians were further identified by ethnicity of Chinese, Filipino, Japanese, Korean, Vietnamese, other Southeast Asian, and South Asian. Age-adjusted incidence rates (AAIRs per 100,000) by race/ethnicity and time period were calculated. The average annual percent changes (AAPC) in incidence rates between 1988 and 2015 were estimated using Joinpoint Regression. Results: The AAIRs of pancreatic cancer varied significantly across the aggregated racial/ethnic groups. The AAIRs ranged from the highest of 16.0 (95% CI: 15.6 to 16.4) in NH blacks to 11.5 (95% CI: 11.5 to 11.6) in NH whites, 10.7 (95% CI: 10.5 to 10.8) in Hispanics, 9.2 (95% CI: 9.0 to 9.4) in NH APIs, and to the lowest of 8.0 (95% CI: 7.2 to 8.9) in NH AIAN. Despite the relatively low rate in the NH APIs, the rates across Asian subgroups varied significantly, with rates as high as NH whites observed in Japanese (AAIR: 11.5; 95% CI: 10.9 to 12.1) and Korean (AAIR: 11.0; 95% CI: 10.3 to 11.7) to the low rate in South Asians (AAIR: 6.6; 95% CI: 5.9 to 7.4). A small but significant decline in pancreatic cancer incidence rates was observed among NH blacks (AAPC: -0.5; 95% CI: -0.8 to -0.2). NH AIANs experienced the most sizable and significant surge in pancreatic cancer incidence (AAPC: 4.5; 95% CI: 2.4 to 6.6), followed by South Asians (AAPC: 2.4; 95% CI: 0.3 to 4.5), Koreans (AAPC: 1.6; 95% CI: 0.5 to 2.7), and NH whites (AAPC: 0.4; 95% CI: 0.2 to 0.6). Conclusions: Our findings provide the most updated trends in pancreatic cancer incidence across multiple and disaggregated racial/ethnic groups in California. We showed significant heterogeneity of pancreatic cancer incidence in Asian Americans, with high-risk Japanese and Korean populations. These results fill a gap regarding pancreatic cancer burden in Asian Americans and underscore the importance of disaggregating ethnic populations in cancer research. Citation Format: Lihua Liu, Juanjuan Zhang, Dennis Deapen, Mariana Stern, Andrea Sipin, Veronica Wendy Setiawan. Racial/ethnic disparities in pancreatic cancer incidence rates and temporal trends in California (1988-2015) [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr C085.

Resource Utilization and Costs in the Care of Patients with Hematologic Malignancies

Background: Patients with hematologic malignancies (HMs) make up a significant portion of the healthcare cost burden. To help understand how patients are managed and to guide allocation of public funds, the utilization of healthcare resources in various phases of disease should be investigated. We examined resource utilization and cost patterns during the pre-diagnosis, treatment, follow-up, and end-of-life phases of patients with specific HMs. Methods: We used the Cancer Care Ontario to identify patients with a diagnosis of HMs, including diffuse large B-cell lymphoma (DLBCL) and Hodgkin lymphoma (HL) between 2006 and 2014. For phase-based analysis, we defined the following four phases of care: Pre-diagnosis: 90 days prior to Ontario Cancer Registry (OCR) diagnosis to index dateInitial treatment: index date to 6 months from OCR diagnosis dateFollow-up: from end of treatment phase to beginning of end-of-life (or completion of cohort follow-up)End-of-life: last 6 months of life We characterized resource utilization by determining overall and disaggregated health system costs for DLBCL and HL. Descriptive statistics were used to characterize the population, with continuous variables presented as medians (with interquartile ranges); costs ($CAN 2014) and resource utilization were normalized to 30-days and presented as means +/- standard deviation. Results: A total of 35,556 patients were diagnosed between 2006 and 2014 with a HM. Of these, 15% were diagnosed with DLBCL, and 7% with HL. There were 5,392 patients diagnosed with DLBCL, [53% male, median age 64 years (IQR 53-74)]. The median follow-up was 1,903 days (IQR 1,194-2,882) from diagnosis, and the median age at death was 72 (IQR 61-82). Mean overall 30-day cost was $1,175 (±2,267) in pre-diagnosis; $9,166 (±5,581) during initial treatment; $1,462 (±2,756) during follow-up; and $7,965 (±7,104) during end-of-life. Mean 30-day cost of cancer medication was 80.20 (±454.98) in pre-diagnosis, 3,115.03 (±1,235.76) during initial treatment, 232.48 (±711.93) during follow-up, and 304.63 (±694.28) during end-of-life. There were 2,367 patients with HL [53% male, median age 37 years (IQR 26-53)]. The median follow-up was 2,419 days (IQR 1,581-3,318) from diagnosis, and the median age at death was 62 (IQR 44-75). In HL, mean overall 30-day cost in Canadian dollars was $813 (±1,541) in pre-diagnosis, $5,473 (±3,485) during initial treatment, $895 (±1,619) during follow-up, and $8,678 (±9,086) during end-of-life. Mean 30-day cost of cancer medication was 91.08 (±368.25) in pre-diagnosis, 1,131.91 (±765.85) during initial treatment, 102.86 (±341.69) during follow-up, and 371.05 (±1,612.73) during end-of-life. Conclusions: Total cost per 30 days was highest in the initial treatment phase and at end-of-life. The highest costs are generally associated with inpatient care across all phases. This suggests that patients require significantly more health care resources during end-of-life and while on active treatment. Results can be used to guide allocation of health care dollars to ensure appropriate patient care throughout a patient's cancer journey. Disclosures No relevant conflicts of interest to declare.