Population-based Case-control Study Research Articles

Tattoos as a new risk factor for malignant lymphoma: a population-based case-control study

Tattoos as a new risk factor for malignant lymphoma: a population-based case-control study

Does tattoo exposure increase the risk of skin cancer? A population-based case-control study

Does tattoo exposure increase the risk of skin cancer? A population-based case-control study

Associations between dietary antioxidant vitamins and risk of glioma: an updated systematic review and meta-analysis of observational studies

BackgroundEpidemiological studies investigating the potential associations between antioxidant vitamins intake and risk of glioma have yielded inconsistent results. To address this, we carried out a systematic review and updated meta-analysis to explore the relationship between dietary antioxidant vitamins intake and risk of glioma.MethodsWe comprehensively searched electronic databases including PubMed, Web of Science, Embase, Scopus, China National Knowledge Infrastructure (CNKI) and Wan fang Data from their inception to March 2024. We employed fixed-effects or random-effects models to estimate the pooled relative risks (RRs) and 95% confidence intervals (CIs) for the associations between dietary antioxidant vitamins intake and risk of glioma. Publication bias was assessed through the visual inspection of the funnel plots and quantified by the Begg's and Egger's tests. Heterogeneity across studies was assessed using the Cochran's Q test and I-square (I2). Additionally, subgroup and sensitivity analyses were performed to explore potential sources of heterogeneity and evaluate the robustness of the results.ResultsOverall, a total of 15 articles involving 3,608 glioma cases and 771,930 participants were included in the final analysis. The pooled analyses revealed that the highest intake of vitamin C significantly reduced the risk of glioma (RR = 0.78; 95%CI: 0.63–0.96; P = 0.022), compared to the lowest intake. However, no significant associations were observed between vitamin A and vitamin E intake and the risk of glioma (P>0.05). Subgroup analyses revealed the inverse association between vitamin C intake and risk of glioma in the population-based case-control studies (RR = 0.82; 95%CI: 0.68–1.00, P = 0.049) and study quality <7(RR = 0.52, 95%CI: 0.29–0.92, P = 0.025).ConclusionOur findings show that higher intake of vitamin C is strongly associated with a reduced risk of glioma, although a dose-response relationship was not evident. Future large-scale prospective studies are warranted to confirm these findings.

Acculturating to multiculturalism: a new dimension of dietary acculturation among Asian American, Native Hawaiian, and Pacific Islander women in the San Francisco Bay Area, USA

BackgroundDietary acculturation is the process by which diet and dietary practises from the environment of origin are retained or changed and/or those prevalent in a new environment are adopted. Despite rapid population growth the U.S., knowledge gaps exist on characterising dietary acculturation among Asian American, Native Hawaiian, and Pacific Islander communities (AANHPI). This study characterise dietary patterns in a sample representative of AANHPI on key demographic characteristics.MethodsData were from a 2013–2014 population-based case-control study in the San Francisco Bay Area, U.S. Survey items were adapted from dietary acculturation scales developed for AANHPI populations. Validated measures assessed social capital, social standing, discrimination and immigration experiences. A principal components factor analysis was conducted to characterise dietary patterns of acculturation.ResultsThree dietary patterns were identified: “Asian,” “Western,” and a distinct “Multicultural” factor. Respondents reporting a high-Asian diet tended to also report smaller social networks, higher levels of stress, and, among those born outside of the U.S., an educational standing that was better before immigration. Respondents reporting a high-Western diet tended to also report the highest level of discrimination. Those reporting a high-Multicultural diet tended to report higher neighbourhood collective efficacy.ConclusionsThe finding of a distinct “Multicultural” factor beyond the typical “Asian” and “Western” factors may reflect the multidirectional relationships between culture, diet, and dietary behavior, in which origin and destination cultures interact in complex ways and where foods from multiple ethnicities intermix.

Uranium in Drinking Water and Bladder Cancer: A Case-Control Study in Michigan.

Uranium is naturally occurring in groundwater used for drinking; however, health risks from naturally occurring concentrations are uncertain. Uranium can cause both radiological and chemical toxicity following ingestion. Bladder and kidneys receive a dose when uranium is excreted into the urine. Investigate the association between uranium in drinking water and bladder cancer risk in a case-control study. A population-based bladder cancer case-control study was conducted in 11 counties of southeastern Michigan. A total of 411 cases and 566 controls provided drinking water and toenail samples and answered questions about lifestyle and residential history. Uranium was measured in drinking water and toenails, and its association with bladder cancer was assessed via unconditional logistic regression models. Median uranium concentration in water was 0.12 μg L-1, with a maximum of 4.99 μg L-1, and median uranium concentration in toenails was 0.0031 μg g-1. In adjusted regression models, there was a suggestion of a protective effect among those exposed to the upper quartile of uranium in drinking water (HR = 0.64, 95% CI: 0.43, 0.96) and toenails (HR 0.66; 95% CI 0.45, 0.96) compared to those in the lowest quartile. Our objective is to investigate additional adjustment of drinking water source at home residence at time of recruitment to address potential selection bias and confounding attenuated results toward the null for drinking water uranium (HR = 0.68, 95% CI: 0.44, 1.05) and toenail uranium (HR = 0.80, 95% CI: 0.53, 1.20). This case-control study showed no increased risk of bladder cancer associated with uranium found in drinking water or toenails.

2D:4D digit ratio as a potential marker for prostate cancer risk

BackgroundThe second-to-fourth digit ratio (2D:4D) is thought to reflect prenatal exposure to sex steroids. We investigated the relationship between 2D:4D and odds of prostate cancer. MethodData were collected in PROtEuS, a population-based case-control study conducted in Montréal, Canada (2005–2012), including 1931 incident prostate cancer cases aged < 76 years and 1994 population controls. In-person interviews elicited information on potential risk factors. Digit lengths were measured by interviewers applying a standard protocol. Odds ratios (OR) and 95 % confidence intervals (CI) were estimated using unconditional logistic regression adjusting for potential confounders. ResultsThe OR of prostate cancer for a standard deviation increase in 2D:4D was 0.91 (95 % CI: 0.85–0.98). For less and more aggressive cancers, ORs were 0.93 (95 % CI: 0.87–1.00) and 0.85 (95 % CI: 0.77–0.93), respectively. There was an interaction with ancestry (p=0.04), whereas the OR among men of African descent was 1.23 (95 % CI: 0.96–1.57, based on 128 cases). ConclusionFindings suggest an inverse association between 2D:4D and odds of overall prostate cancer, more pronounced for aggressive cancers. This supports the notion that high levels of testosterone in utero, estimated by a low 2D:4D ratio, are associated with a higher risk of prostate cancer. Contrastingly, a high digit ratio was associated with greater cancer odds among participants of African descent. Upon replication, 2D:4D could prove to be an easily measured marker of prostate cancer risk.

Interaction between alimentary surgery and risk of dementia: a nationwide population-based case-control study.

Based on current research, it is known that the gastrointestinal tract microbiota and its genome play a crucial role in mental illnesses. Studies indicate a direct correlation between gastrointestinal tract microbiota and the onset of dementia, mediated by metabolic diseases and low-grade inflammation. The association between various gastrointestinal symptoms and neurodegenerative diseases has been recently discussed. However, there is a lack of research regarding the comparative effects of different surgical procedures on neurodegenerative diseases. Therefore, this study primarily focuses on comparing the association between various gastrointestinal surgeries and dementia, aiming to provide guidance for future clinical practice. A nationwide study using the Taiwanese National Health Insurance Research Database included 26 059 patients diagnosed with dementia or Alzheimer's disease and 104 236 controls without diseases. Primary exposures were defined as alimentary surgeries, encompassing cholecystectomy, gastrectomy, bowel resection, and appendectomy. Conditional logistic regression was used to examine the odds ratio and 95% confidence interval for prior alimentary surgery between cases and controls. The results showed that individuals with dementia had a higher rate of gastrectomy. Additionally, individuals with dementia seemed to exhibit a reduced rate of cholecystectomy and appendectomy. Regarding Alzheimer's disease, all four alimentary surgeries showed comparable trends to those observed with dementia. No significant interaction was observed between alimentary surgery and dementia among the four types of surgery evaluated. Our study demonstrates that gastrectomy is associated with an elevated risk of dementia. We aim to uncover more direct evidence in future experiments.

Epidemiology and association of neighbourhood marginalisation on violent knife assaults in Ontario: a population-based case-control study

BackgroundViolent knife assaults (‘stabbings’) are underappreciated as a source of morbidity and mortality. The two objectives of this study were to describe the epidemiology of stabbing injuries in the population...

Proton pump inhibitor use and bone fractures in patients with chronic kidney disease.

Patients with chronic kidney disease (CKD) are at high risk for bone fractures, which are associated with high morbidity and mortality. Proton pump inhibitors (PPI) have been linked to an increased risk for fractures in the general population as well as in patients with need for hemodialysis, but studies in patients with CKD are currently missing. We performed a population-based observational case-control study exploring a sample of patients with CKD derived from the IQVIATM Disease Analyzer database. Patients with and without fractures were matched using the 1:1 nearest neighbor propensity score matching method. To investigate the association between PPI use and fractures, multivariable logistic regression analyses were performed adjusting for confounding factors. In total, 6076 patients with and 6076 patients without fractures were matched and subsequently available for analyses. In the total cohort, PPI use was associated with an increased risk for fractures (OR 1.68; 95% CI 1.55-1.83). This association was noted for nearly all types of fractures. The strongest association between PPI use and fractures was found in patients below the age of 60 with a PPI prescription for longer than two years (OR 6.85, 95% CI 1.85-25.38). The same was true, when analyzing cumulative PPI doses. Here, patients below the age of 60 with a cumulative PPI dose above 16000mg (highest quartile) had the highest risk for fractures (OR 4.62, 95% CI 1.87-11.44). There was no difference between men or women regarding the association between PPI use and fractures. This study provides evidence that PPI use is associated with fractures in patients with CKD. Deprescription of PPI in patients without an indication for treatment could be a modifiable risk factor to reduce fracture risk in this high-risk group.

Social networks, social determinants, and mortality: Western New York Exposures and Breast Cancer study.

There are few studies of social support and other social determinants of health after breast cancer diagnosis and their associations with mortality; results have been inconclusive. Further, it is not known if observed associations are specific to women with breast cancer diagnosis or if associations would be similar among healthy women. Women with incident, pathologically confirmed invasive breast cancer, stage I-IV (n = 1012), and healthy frequency age-matched participants (n = 2036) answered a social support questionnaire in prospective follow-up of a population-based case-control study, the Western New York Exposures and Breast Cancer Study. At interview, all participants were aged 35-79 years and resident of 2 counties in Western New York State. Mortality status was ascertained from the National Death Index. Participants were queried regarding the number of their close friends, frequency of seeing them, household size, household income, and marital status. Hazard ratios (HRs) and 95% confidence intervals (CIs) for breast cancer-specific mortality (breast cancer women only) and all-cause mortality were estimated. Lower household income was associated with higher all-cause mortality among women diagnosed with breast cancer (HR = 2.48, 95% CI = 1.24 to 4.97) and similarly among the healthy women (HR = 2.63, 95% CI = 1.25 to 5.53). Number and frequency of seeing friends, marital status, and household size were not associated with mortality, either among breast cancer patients or among healthy women. Among those diagnosed with breast cancer and healthy women, lower income was associated with more than twice the mortality. Marital status, household size, and number or frequency of meeting friends were not associated with survival.

Breast Cancer is Increased in Women with Primary Ovarian Insufficiency.

DNA damage/repair gene variants are associated with both primary ovarian insufficiency (POI) and cancer risk. We hypothesized that a subset of women with POI and family members would have increased risk for cancer. Case-control population-based study using records from 1995-2022. Two major Utah academic healthcare systems serving 85% of the state. Women with POI (n=613) were identified using ICD codes and reviewed for accuracy. Relatives were linked using the Utah Population Database. Cancer diagnoses were identified using the Utah Cancer Registry. The relative risk of cancer in women with POI and relatives was estimated by comparison to population rates. Whole genome sequencing was performed on a subset of women. Breast cancer was increased in women with POI (OR [95%CI] 2.20 [1.30, 3.47]; p=0.0023) and there was a nominally significant increase in ovarian cancer. Probands with POI were 36.5±4.3 years and 59.5±12.7 years when diagnosed with POI and cancer, respectively. Causal and candidate gene variants for cancer and POI were identified.Among second-degree relatives of these women, there was an increased risk of breast (1.28 [1.08, 1.52]; p=0.0078) and colon cancer (1.50 [1.14, 1.94]; p=0.0036). Prostate cancer was increased in first- (1.64 [1.18, 2.23]; p=0.0026), second- (1.54 [1.32, 1.79]; p<0.001), and third-degree relatives (1.33 [1.20, 1.48]; p<0.001). Data suggest common genetic risk for POI and reproductive cancers. Tools are needed to predict cancer risk in women with POI and potentially to counsel about risks of hormone replacement therapy.

COVID-19 infection during pregnancy and risk of early and late spontaneous miscarriages: A matched case-control population-based study.

To evaluate the effect of COVID-19 during the first trimester on the rate of first- and second-trimester miscarriages. Secondary aims include the effect on stillbirths and the correlation between symptom severity and pregnancy outcomes. A retrospective matched case-control population-based study extracted data from electronic medical records of a nationwide database of the second largest healthcare organization that provides medical services to over 2 000 000 patients in Israel. Pregnancy outcomes in COVID-19-positive pregnant patients in 2020 were compared with an age- and gestational-week-matched 1:2 case-control cohort of pre-pandemic pregnant patients that received medical care in 2019. Of 68 485 pregnant women treated in 2020, 2333 were COVID-19-positive during pregnancy: 215 during the first trimester, 791 during the second trimester, and 1327 during the third trimester. We compared these data with the control cohort of 4580 pre-pandemic pregnant patients. The rate of spontaneous miscarriage was significantly higher 146/2187 (6.3%) in COVID-19-positive patients versus 214/4580 (4.7%), (P < 0.01, odds ratio 1.34, 95% confidence interval 1.094-1.691). Most miscarriages occurred during the first trimester in both groups, yet the rates were significantly higher in the study group (5.4% vs 3.8%, P < 0.01). There was no association between COVID-19 severity and miscarriage risk. COVID-19 diagnosis during early pregnancy increased the rate of spontaneous miscarriage in our cohort compared with an age- and gestational-week-matched pre-pandemic control group.

Relation between the Dantu blood group variant and bacteraemia in Kenyan children: a population-based case control study.

The Dantu blood group variant protects against P. falciparum infections but its wider consequences have not been previously explored. Here, we investigate the impact of Dantu on susceptibility to bacteraemia. We conducted a case-control study in children presenting with community-acquired bacteraemia to Kilifi County Hospital in Kenya between 1998 and 2010. We used logistic regression to test for associations between the Dantu marker SNP rs186873296 A>G and both all-cause and pathogen-specific bacteraemia under an additive model. We used date of admission as a proxy measure of malaria transmission intensity, given known differences in malaria prevalence over the course of the study. Dantu was associated with protection from all-cause bacteraemia (OR=0.81, p=0.014), the association being greatest in homozygotes (OR=0.30, p=0.013). This protection was shared across the major bacterial pathogens but, notably, was only significant during the era of high malaria-transmission pre-2003 (OR=0.79, p=0.023). Consistent with previous studies showing the indirect impact on bacteraemia risk of other malaria-associated red cell variants, our study also shows that Dantu is protective against bacteraemia via its effect on malaria risk. Dantu does not appear to be under balancing selection through an increased risk of bacterial infections.

Residential exposure to ambient fine particulate matter (PM2.5) and nitrogen dioxide (NO2) and incident breast cancer among young women in Ontario, Canada

BackgroundAir pollution has been classified as a human carcinogen based largely on findings for respiratory cancers. Emerging, but limited, evidence suggests that it increases the risk of breast cancer, particularly among younger women. We characterized associations between residential exposure to ambient fine particulate matter (PM2.5) and nitrogen dioxide (NO2) and breast cancer. Analyses were performed using data collected in the Ontario Environmental Health Study (OEHS). MethodsThe OEHS, a population-based case-control study, identified incident cases of breast cancer in Ontario, Canada among women aged 18–45 between 2013 and 2015. A total of 465 pathologically confirmed primary breast cancer cases were identified from the Ontario Cancer Registry, while 242 population-based controls were recruited using random-digit dialing. Self-reported questionnaires were used to collect risk factor data and residential histories. Land-use regression and remote-sensing estimates of NO2 and PM2.5, respectively, were assigned to the residential addresses at interview, five years earlier, and at menarche. Logistic regression was used to estimate odds ratios (OR) and their 95 % confidence intervals (CI) in relation to an interquartile range (IQR) increase in air pollution, adjusting for possible confounders. ResultsPM2.5 and NO2 were positively correlated with each other (r = 0.57). An IQR increase of PM2.5 (1.9 µg/m3) and NO2 (6.6 ppb) at interview residence were associated with higher odds of breast cancer and the adjusted ORs and 95 % CIs were 1.37 (95 % CI = 0.98–1.91) and 2.33 (95 % CI = 1.53–3.53), respectively. An increased odds of breast cancer was observed with an IQR increase in NO2 at residence five years earlier (OR = 2.16, 95 % CI: 1.41–3.31), while no association was observed with PM2.5 (OR = 0.96, 95 % CI 0.64–1.42). ConclusionsOur findings support the hypothesis that exposure to ambient air pollution, especially those from traffic sources (i.e., NO2), increases the risk of breast cancer in young women.

Adolescent sleep patterns, genetic predisposition, and risk of multiple sclerosis.

Shift work, insufficient sleep, and poor sleep quality at young age have been associated with increased risk of multiple sclerosis (MS). This study aimed to investigate the potential interaction between aspects of inadequate sleep (short sleep, phase shift, and poor sleep quality) during adolescence and HLA-DRB1*15:01 in relation to MS risk. We used a Swedish population-based case-control study (1253 cases and 1766 controls). Subjects with different sleep patterns during adolescence and HLA-DRB1*15:01 status were compared regarding MS risk by calculating odds ratios (OR) with 95% confidence intervals (CI) using logistic regression models. Additive interaction between aspects of inadequate sleep and HLA-DRB1*15:01 status was assessed by calculating the attributable proportion due to interaction (AP) with 95% CI. Short sleep duration (<7 hours/night) during adolescence acted synergistically with HLA-DRB1*15:01, increasing the risk of MS (AP 0.38, 95% CI 0.01-0.75, p=0.04). Similarly, subjective low sleep quality during adolescence interacted with HLA-DRB1*15:01 regarding risk of MS (AP 0.30, 95% CI 0.06-0.56, p=0.03), whereas phase shift did not significantly influence the risk of the disease, irrespective of HLA-DRB1*15:01 status. Our findings underscore the importance of addressing inadequate sleep during adolescence, particularly in the context of the HLA-DRB1*15:01 allele, as it appears to amplify the risk of subsequently developing MS.

Orofacial Clefts and Maternal Risk Factors: A Population-Based Case-Control Study.

Orofacial clefts (OFCs) are some of the most common congenital anomalies worldwide. The aim of this case-control study was to evaluate the association of OFCs with selected maternal characteristics. Data on isolated non-syndromic cases of OFCs were extracted from the population-based registry of congenital anomalies of Tuscany. A sample of live-born infants without any congenital anomaly was used as the control group. We investigated the association with sex and some maternal characteristics: age, body mass index, smoking, and education. Adjusted odds ratios (OR) were calculated using a logistic regression model. Analyses were performed for the total OFCs and separately for cleft lip (CL) and cleft palate (CP). Data on 219 cases and 37,988 controls were analyzed. A higher proportion of males (57.9%) was observed, particularly for CL. A decreasing trend among the maternal age classes was observed (OR:0.81 (95%CI 0.70-0.94)). Underweight mothers had a higher prevalence of OFCs, in particular for CL (OR:1.88 (95%CI 1.08-3.26)). We found an association of OFCs with lower maternal age. The association with maternal age remains controversial and further epidemiological evidence is needed through multicenter studies. We observed that CL was more common in underweight mothers, suggesting actions of primary prevention.

Inflammatory eye disease is a risk factor for future microscopic colitis: A nationwide population-based matched case control study.

Microscopic colitis (MC) is an inflammatory disorder of the colon. To date, the relationship between inflammatory eye diseases and MC is unclear. To assess whether inflammatory eye disease (iridocyclitis and episcleritis) is a risk factor for MC. We conducted a nationwide matched case control study in Sweden leveraging the ESPRESSO-study (a Swedish database containing data on all biopsies from the gastrointestinal tract from 1965 to 2017). In total, we identified 14,338 patients with biopsy-verified MC (diagnosed from 1981 to 2017). Patients with MC were matched (by age, sex, county and year of birth) with 68,753 controls from the general population and the occurrence of preceding inflammatory eye diseases (defined as diagnosis of episcleritis or iridocyclitis) in the two groups was compared. Multivariable adjusted odds ratios (aORs) were calculated using conditional logistic regression conditioned on the matching variables. A majority of patients with MC were women (71.9%) and the median age at MC diagnosis was 63.3years (interquartile range (IQR)=50.7-72.6). Some 225 (1.6%) MC patients had an earlier record of inflammatory eye disease compared with 614 (0.9%) in controls. These figures corresponded to an aOR of 1.77 (95% CI=1.52-2.07) for inflammatory eye diseases in patients with MC. Compared to siblings, the aOR for previous inflammatory eye diseases in MC was 1.52 (95% CI=1.17-1.98) and patients treated with budesonide, as a proxy for clinically significant disease, had a somewhat higher aOR for previous inflammatory eye diseases. Inflammatory eye diseases are more common in patients subsequently being diagnosed with MC. Our findings highlight that these conditions may have shared causes and inflammatory pathways and are of clinical interest to gastroenterologists, ophthalmologists and general practitioners.

Breast cancer immunophenotypes and serum organochlorine pesticides in Mexican women: Mixture exposure approach

Epidemiological studies on associations between breast cancer (BC) and organochlorine pesticides (OCP) are inconclusive. The majority of studies have evaluated the effect of single compounds without considering multiple OCP exposures and immunophenotypes of BC. We aimed to evaluate the association between BC immunophenotypes and serum OCP mixtures, and identify the main contributors within mixtures. We included 767 histopathologically confirmed incident BC cases and 908 controls from a population-based case-control study conducted from 2007 to 2011 in Northern Mexico. We obtained direct information about sociodemographic, lifestyle and reproductive characteristics. We collected data from clinical records about hormonal receptors (HR) and epidermal growth factor receptor 2 (HER2) expressions. Immunophenotypes were determined as HR+/HER2−, HER2+ or HR−/HER2−. We quantified OCP and metabolites by gas chromatography using an electron capture micro detector. We used Weighted Quantile Sum regression to assess the association of BC and exposure to multiple OCP, and their contribution within the mixture. We found a positive adjusted association between BC and an OCP mixture (OR: 3.48, 95%CI: 2.58, 4.69), whose primary contribution arose from the isomers of hexachlorocyclohexane and endosulfan, as well as endosulfan sulfate. We also identified a mixture negatively associated (OR: 0.13, 95%CI: 0.08, 0.20), characterized by p,p'-DDT and chlordane metabolites. All these associations remained regardless BC immunophenotypes. This is the first epidemiological report that identified serum OCP mixtures associated with BC immunophenotypes. Due to OCP ubiquity, biomagnification, and continuous exposure, they constitute a global problem of persistent exposure that might be related to BC risk.

Newborn readmissions and virtual primary care delivery: a population-based case-control study

BackgroundEarly post-discharge assessments for newborns are recommended. Virtual care has become more prevalent during the pandemic, providing an opportunity to better understand its impact on the quality of post-discharge newborn care. The objective of this study was to understand whether primary care visit modality (in-person vs. virtual) is associated with early newborn hospital readmissions and emergency department (ED) visits.MethodsWe conducted a population-based, case-control study using linked health administrative databases between September 1, 2020 and March 31, 2022 in Ontario, Canada. We compared the modality of primary care visits among cases (hospital readmission within 14 days of life) and controls (newborns without a readmission), matched on infant sex, gestational age, and maternal parity. We included an alternative definition of cases as a composite of either a newborn hospital readmission or emergency department (ED) visit or in-hospital death within the first 14 days of life. Conditional logistic regression models were used to model odds ratios (ORs), comparing those exposed to a virtual visit versus in-person visit, adjusting for infant birth weight, birth hospitalization length of stay, neighbourhood level material deprivation, rurality and presence of active maternal comorbidities.ResultsAmong 73,324 eligible newborns, 2,220 experienced a hospital readmission within 14 days of life and were matched to 8,880 controls. Jaundice was the primary reason for readmission (75% of readmissions). Compared to newborns who were seen in-person post-discharge, newborns who were seen virtually had higher odds of hospital readmission (adjusted odds ratio [aOR] 1.41 (95% CI 1.09, 1.83); the magnitude of effect was not different using the composite outcome (aOR 1.35, 95% CI 1.05, 1.75).ConclusionsNewborns who receive a virtual post-discharge visit are more likely than those who receive an in-person visit to require hospital readmission.

Predicting Out-of-Hospital Cardiac Arrest in the General Population Using Electronic Health Records.

The majority of out-of-hospital cardiac arrests (OHCAs) occur among individuals in the general population, for whom there is no established strategy to identify risk. In this study, we assess the use of electronic health record (EHR) data to identify OHCA in the general population and define salient factors contributing to OHCA risk. The analytical cohort included 2366 individuals with OHCA and 23 660 age- and sex-matched controls receiving health care at the University of Washington. Comorbidities, electrocardiographic measures, vital signs, and medication prescription were abstracted from the EHR. The primary outcome was OHCA. Secondary outcomes included shockable and nonshockable OHCA. Model performance including area under the receiver operating characteristic curve and positive predictive value were assessed and adjusted for observed rate of OHCA across the health system. There were significant differences in demographic characteristics, vital signs, electrocardiographic measures, comorbidities, and medication distribution between individuals with OHCA and controls. In external validation, discrimination in machine learning models (area under the receiver operating characteristic curve 0.80-0.85) was superior to a baseline model with conventional cardiovascular risk factors (area under the receiver operating characteristic curve 0.66). At a specificity threshold of 99%, correcting for baseline OHCA incidence across the health system, positive predictive value was 2.5% to 3.1% in machine learning models compared with 0.8% for the baseline model. Longer corrected QT interval, substance abuse disorder, fluid and electrolyte disorder, alcohol abuse, and higher heart rate were identified as salient predictors of OHCA risk across all machine learning models. Established cardiovascular risk factors retained predictive importance for shockable OHCA, but demographic characteristics (minority race, single marital status) and noncardiovascular comorbidities (substance abuse disorder) also contributed to risk prediction. For nonshockable OHCA, a range of salient predictors, including comorbidities, habits, vital signs, demographic characteristics, and electrocardiographic measures, were identified. In a population-based case-control study, machine learning models incorporating readily available EHR data showed reasonable discrimination and risk enrichment for OHCA in the general population. Salient factors associated with OCHA risk were myriad across the cardiovascular and noncardiovascular spectrum. Public health and tailored strategies for OHCA prediction and prevention will require incorporation of this complexity.