Sex-stratified analyses of comorbidities associated with an inpatient delirium diagnosis using real world data

Sex-stratified analyses of comorbidities associated with an inpatient delirium diagnosis using real world data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Sex-stratified analyses of comorbidities associated with an inpatient delirium diagnosis using real world data Marina Sirota, Lay Kodama, Sarah Woldemariam, Alice Tang, Yaqiao Li, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4765249/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Jul, 2025 Read the published version in Communications Medicine → Version 1 posted You are reading this latest preprint version Abstract Delirium is a detrimental mental condition often seen in older, hospitalized patients and is currently hard to predict. In this study, we leverage electronic health records (EHR) to identify 7,492 UCSF patients and 19,417 UC health system patients with an inpatient delirium diagnosis and the same number of control patients without delirium. We found significant associations between comorbidities or laboratory values and an inpatient delirium diagnosis, including metabolic abnormalities and psychiatric diagnoses. Some associations were sex-specific, including dementia subtypes and infections. We further explored the associations with anemia and bipolar disorder by conducting longitudinal analyses from the time of first diagnosis to development of delirium, demonstrating a significant relationship across time. Finally, we show that an inpatient delirium diagnosis leads to increased risk of mortality. These results demonstrate the powerful application of the EHR to shed insights into prior diagnoses and laboratory values that could help predict development of inpatient delirium and the importance of sex when making these assessments. Health sciences/Diseases/Neurological disorders/Disorders of consciousness Health sciences/Diseases/Neurological disorders/Dementia Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Main Delirium is a clinical diagnosis defined as fluctuating disturbances in attention, awareness, and cognition that develops over a short period of time and is highly prevalent among older inpatient populations, with an estimated prevalence of 23% in hospitalized older adults 1 , 2 . Both long- and short-term outcomes of delirium are detrimental to patients, distressing to caregivers, and a burden on the healthcare system 3 – 5 . Notably, patients who are diagnosed with delirium have an increased risk of long-term mortality. It is unclear, however, whether delirium is an independent risk factor for mortality versus other confounding factors are at play such as overall baseline disease burden 3 . Medications for treating delirium are largely for symptomatic management and have not been shown to have clinical benefits and in fact may lead to worse outcomes, especially for older patients 6 . Prevention is the most effective strategy and often involve non-pharmacological interventions, such as reorientation and cognitive stimulation 7 , making it imperative to predict which patients may develop delirium using predictive tools to focus prevention efforts for high-risk patients. Though prevalent, delirium remains challenging to study and predict given its heterogenous clinical nature and diverse risk factors 2 . Current prediction tools have variable predictive capabilities, with one systematic review finding an area under the receiver operating curve range from 0.52 to 0.94 8 . These models have other limitations including validation in small sample sizes or data collected from one institution that may not be generalizable 9 – 11 . These models incorporate well-studied risk factors such as existing cognitive impairment and severity of chronic illnesses, but less apparent risk factors also need to be identified to improve these predictive tools 8 , 12 , 13 . Importantly, sex is a major risk modifier for many neurological diseases, with some evidence of sex-differences in delirium 14 – 16 . For instance, one study found hypoactive delirium to be more common in female compared to male patients 15 . Though clinical sex differences have been well-documented in dementia and cognitive impairment -- major risk factors for delirium -- whether sex differences exist in delirium remains largely unstudied due to the lack of sex-stratified studies. One strategy to overcome these limitations is to employ large-scale, comprehensive real-world data from electronic health records (EHR) combined with robust computational approaches. Data from the EHR have been used to better understand and phenotype complex diseases such as Alzheimer’s disease, type 2 diabetes, and preterm birth 17 – 19 . Such deep phenotyping studies can shed insights into clinical risk factors and subtypes of disease as well as point to potential new biological pathways that may be involved in disease. In this study, we leveraged EHRs from two databases across California to identify differential comorbidities and laboratory test results prior to a patient’s first inpatient admission for delirium, which could serve as potential factors in identifying patients at risk for the condition. Importantly, we also conducted a sex-stratified analysis to identify differences by documented sex in the association between comorbidities or laboratory findings with a delirium diagnosis. We further analyzed specific comorbidities using a longitudinal approach to predict the length of time from the diagnosis of a comorbidity and admission for delirium, as an alternative, complementary approach to validate the association study and also provide granularity into the time course from comorbidity diagnosis to delirium diagnosis. Lastly, we conducted a longitudinal analysis to understand the relationship between first inpatient delirium diagnosis and mortality outcomes. Results Identification of patients with a diagnosis of delirium and their matched controls We identified 7,492 patients with an inpatient delirium diagnosis and 7,492 propensity-score (PS)-matched control patients within the University of California, San Francisco (UCSF) de-identified electronic health record (EHR) database (~ 5 million patients total). Propensity score matching was used to create a matched case-control study as individual covariates may not be well-matched but the “propensity” to be either a case or control is well-matched, as illustrated in prior publications justifying this method 20 , 21 . Control patients were matched on the following demographic and inpatient visit features: age at admission, patient-identified race, sex, death during admission, length of the inpatient visit of interest in days, years of available EHR data, total number of inpatient visits prior to the visit of interest, total number of comorbidities prior to the visit of interest, and whether the visit of interest was in the ICU setting. The “visit of interest” corresponded to the visit where an inpatient delirium diagnosis was made for the delirium group or a randomly selected inpatient visit for the control group (Fig. 1 ). Similarly, a separate cohort of 19,417 patients with an inpatient delirium diagnosis and 19,417 PS-matched control cohort were identified from the UC-Wide EHR database (~ 8.6 million patients total, data from UC Davis, UC Los Angeles, UC Irvine, UC San Diego) with an additional matching criterion that included UC location (Fig. 1 ). These covariates were chosen based on prior EHR-based studies using PS-matched cohort selections with inclusion of additional covariates to control for health status and healthcare utilization frequency using indirect measures such as the total number of inpatient visits and comorbidities prior to the visit of interest and years of available EHR data 17 , 22 – 24 . Because patients who develop delirium tend to be at baseline more ill than those who do not develop delirium, we wanted to find a comparable control group that had similar baseline disease burden and identify what specific comorbidities are differentially enriched between patients who are similarly "ill", but one group develops delirium and the other group does not. Table 1. Demographic information of matched cohorts. Table of demographic information for patient cohorts identified in UCSF (left) and UC-wide (right) data. Chi-squared test used for categorical measures. Student’s t-test used for continuous measures. SMD = standardized mean difference. Post-matching analysis showed adequate matching of covariates with similar PS distributions between the groups ( Extended Data Fig. 1 a,c) and absolute standardized mean differences less than 0.1, except for inpatient stay length which had a wide distribution in both databases ( Extended Data Fig. 1 b,d). Detailed demographic and inpatient visit features for the cohorts generated are shown in Table 1 . We defined inpatient delirium broadly using the Observational Medical Outcomes Partnership (OMOP) concept ID 373995 (corresponding to “Delirium”) and excluded diagnoses that had specific causes of delirium in the diagnosis name (such as alcohol-induced delirium or other substance-induced delirium). These exclusion criteria were used to focus on cases where a delirium diagnosis was made but the cause of the delirium was not readily known. Using this definition, we were able to capture 84% and 89% of all delirium-related visits in the UCSF and UC-wide databases, respectively, with delirium prevalence within the wide range of published estimates for patients 65 years and older (9.5% mean, 8.9% SD in UCSF data; 3.5% mean, 1.9% SD in UC-wide data) ( Extended Data Fig. 2 a,b). We used the Richmond Agitation Sedation Scale (RASS) rating to corroborate the delirium diagnosis in the UC-wide data. RASS is a 10-point rating scale ranging from − 5 to 5, with negative numbers corresponding to the level of sedation and positive numbers corresponding to the level of agitation. Though RASS is a more general measure of a patient’s level of sedation, validated mostly in the ICU setting, it has also been implemented more widely in the inpatient setting for detection of delirium 25 , 26 . Less than half of patients had documented RASS ratings during their visit of interest. Of the available ratings, patients with delirium had a statistically significant increase in the mean RASS rating compared to controls ( Extended Data Fig. 2 c), suggesting patients with a delirium diagnosis are overall more agitated during the visit and also have greater variability in their RASS ratings (data not shown), suggesting a waxing and waning nature to their clinical status. Patients with delirium are more likely to be diagnosed with diseases of the nervous system, mental health, metabolic disorders, and infections compared to control patients To understand potential risk factors associated with an inpatient delirium diagnosis, we first collected all first-time diagnoses made during visits prior to the inpatient admission of interest. Low-dimensional Uniform Manifold Approximation and Projection (UMAP) representation of all non-delirium diagnoses (19,590 features, SNOMED concept IDs) shows a statistically significant separation of patients with a delirium diagnosis versus matched controls by two-sided Mann-Whitney U test (UMAP 1, p-value < 2.2 e-16; UMAP 2 p-value 0.0086; Fig. 2 a). Differential association analyses of these comorbidities using Fisher’s exact test showed enrichment of distinct comorbidities for patients with delirium compared to control patients, with 101 diagnoses significantly enriched in patients with delirium versus 108 in controls, out of 19,583 diagnoses tested (Fig. 2 b). Control patients had enrichment of diagnoses largely related to pregnancy and other health statuses as well as age-related musculoskeletal diagnoses such as pain in joints and osteoarthritis and skin findings such as melanocytic nevus (Fig. 2 b,c, and Supplementary Table 1 ). Meanwhile, patients with delirium had enrichment of diagnoses related to diseases of the nervous system, including epilepsy and seizures, mental health and behavioral disorders, and acute diagnoses such as metabolic diseases and infections (Fig. 2 b,c, and Supplementary Table 1 ). Similar diagnostic associations were seen in the UC-wide cohort based on a hypergeometric test (p-value 1.2 e-94; Fig. 2 d-f, Extended Data Fig. 3 , Supplementary Table 2, and Supplementary Table 3 ). Categorizations of these overlapping diagnoses between the two databases by ICD10 diagnostic blocks showed enrichment of diagnoses related to certain disease categories in patients with delirium versus control patients, including blood-related diseases such as anemia, diseases of the genitourinary system such as urinary tract infections, diseases of the nervous system such as epilepsy, metabolic diseases such as hyponatremia, and mental health-related disorders such as bipolar disorder ( Extended Data Fig. 3 d and Supplementary Table 3 ). These findings are largely consistent with previously-identified risk factors for delirium 2 . Differential laboratory findings corroborate differential associations between comorbidities and delirium We also conducted enrichment analysis for mean laboratory values and vital signs collected before the inpatient admission of interest. Patients with a delirium diagnosis had significantly higher mean values of certain liver function tests, such as alkaline phosphatase and aspartate transferase, compared to controls in both UCSF and UC-wide datasets, suggesting potential liver dysfunction in these patients (Fig. 3 a). Elevated mean vital signs included heart rate and respiratory rate ( Fig. 3 a). Meanwhile, glomerular filtration rate and urine creatinine levels were decreased in patients with delirium compared to controls, consistent with kidney dysfunction (Fig. 3 a ). Hemoglobin, hematocrit, and erythrocyte counts were also significantly decreased in patients with delirium compared to controls, consistent with the association with an anemia diagnosis in these patients (Fig. 3 ). The UC-wide dataset also captured certain clinical test scores, including results from the Patient Health Questionnaire (PHQ)-2 and PHQ-9, consistent with the associations with depression in these patients ( Supplementary Table 4 ). Sex-stratified analysis shows certain infections and dementia subtypes are sex-specific risk factors for delirium To understand whether any of the comorbidities associated with delirium that we identified are sex-specific, we conducted a sex-stratified association analysis using the same cohort identified above ( Extended Data Fig. 4 and Extended Data Fig. 5 ). We identified several diagnoses that were significantly associated with delirium in only females, only males, or in both female and male patients with delirium compared to controls in both UCSF and UC-wide datasets ( Extended Data Fig. 4 c-e and Extended Data Fig. 5 c-e) as well as sex-specific laboratory results ( Extended Data Fig. 4 f and Extended Data Fig. 5 f). Diagnoses common to both male and female patients with delirium in both datasets were largely similar to the non-stratified analysis, including symptoms of delirium such as altered mental status, restlessness and agitation, and hallucinations, as well as known organic causes of delirium and altered mental status such as seizures, cognitive disorders and other mental disorders (Fig. 4 c). Interestingly, female and male patients diagnosed with delirium had sex-specific associations with distinct infections and diseases of the nervous system that were statistically significant in both datasets. For instance, female patients with delirium had associations with encapsulated bacterial infections due to Streptococcal bacteria (OR UCSF 2.38; UC-wide 1.61), Klebsiella pneumoniae (OR UCSF 2.95, UC-wide 2.14), Escherichia coli (OR UCSF 1.45, UC-wide 1.81), and enterococcal bacteria (OR UCSF 1.53, UC-wide 1.79), while males had associations with Clostridioides difficile infections (OR UCSF 3.89, UC-wide 1.55) (Fig. 4 a,b). Sex-specific associations with subtypes of dementia were also seen, where females had significant associations with Alzheimer’s disease (OR UCSF 2.06, UC-wide 3.55) and vascular dementias with behavioral disturbances (OR UCSF 1.89, UC-wide 3.29) while males had associations with Diffuse Lewy Body disease (OR UCSF 1.51, UC-wide 4.22) and diagnoses related to symptoms of the disease such as visual hallucinations, falls, difficulty walking / muscle weakness, dysphagia, insomnia, and constipation (Fig. 4 a,b, Supplementary Table 5, and Supplementary Table 6 ). These sex-specific associations could be because these dementia subtypes are known to have sex-differences in their prevalence already 14 or could point to sex-specific ways in which delirium manifest in different patient populations. Though there were several sex-specific laboratory results found in the UCSF and UC-wide datasets, none were common between the two datasets. Laboratory findings associated with delirium in both male and female patients as well as between both datasets were similar to the results of the non-sex-stratified analysis, including decreased hemoglobin and hematocrit consistent with an anemia diagnosis, decreased GFR consistent with kidney dysfunction, elevated liver function tests such as alkaline phosphatase, and elevation of heart rate (Fig. 3 a, Extended Data Fig. 4 f, and Extended Data Fig. 5 f). Longitudinal analysis from diagnosis of a potential risk factor of delirium to an inpatient delirium diagnosis validates comorbidity association study To further understand several of the comorbidities associated with an inpatient delirium diagnosis identified above, we carried out a longitudinal time-to-event analysis for anemia and bipolar disorder. We chose these diagnoses for analysis given the higher association of blood-related disorders and mental and behavioral disorders in patients with delirium found through our association study (Fig. 2 d). Patients with a first-time diagnosis of anemia and no prior diagnosis of delirium and a matched control group with no diagnosis of anemia and no prior diagnosis of delirium were identified. Control patients were matched on the following parameters: age at the visit of interest, patient-identified race, documented sex, years of available EHR data, total number of inpatient visits prior to the visit of interest, and total number of comorbidities prior to the visit of interest ( Extended Data Fig. 6 ). Events were defined as admission for delirium, death, or loss to follow-up. Kaplan-Meier curve visualization of the data showed stratification by anemia diagnosis status, where those with anemia have increased probability of developing first-time inpatient delirium diagnosis (UCSF 3.4%, UC-wide 1.3% of anemia patients) than those without any anemia diagnosis (UCSF 0.3%, UC-wide 0.3% of control patients) over the course of ~ 30 years in the UCSF data and ~ 11 years in the UC-wide data (Fig. 5 a and Extended Data Fig. 7a ). Cox proportional hazard ratio analysis unadjusted and adjusted for demographics and visit features revealed a significant increased risk of delirium in those with an anemia diagnosis compared to controls (UCSF HR 9.4; 95% CI, 8.1 to 11; UC-wide HR 4.4; 95% CI, 4.1 to 4.7) (Fig. 5 b). A similar analysis was done for patients with and without Bipolar I Disorder (BD1) in UCSF patients or bipolar disorder (unspecified type) in UC-wide patients. We found that a diagnosis of BD also increased risk of developing first-time inpatient delirium diagnosis (UCSF 1.9%, UC-wide 0.7% of BD patients) than those without a BD diagnosis (UCSF 0.01%, UC-wide 0.01% of control patients) with a HR of 27 (95% CI, 9.9 to 74.4) for UCSF patients and HR of 7.8 (95% CI, 6.0 to 10.0) for UC-wide patients over the course of ~ 20 years in the UCSF data and ~ 10 years in the UC-wide data (Fig. 5 c,d, and Extended Data Fig. 7b ). Our association studies also found that certain diagnoses are more enriched in control patients compared to patients with delirium, including diagnoses under the ICD10-CM neoplasm category such as melanocytic nevus (Fig. 2 d and Supplementary Table 3 ). We did a similar time-to-event analysis of patients with and without this diagnosis and found that, indeed, a prior diagnosis of melanocytic nevus modestly decreased the risk of developing delirium at a HR of 0.3 (95% CI, 0.2 to 0.5) for UCSF patients and HR of 0.76 (95% CI, 0.67 to 0.86) for UC-wide patients (Fig. 5 e,f, and Extended Data Fig. 7c ). A single inpatient delirium admission is associated with increased mortality Previous meta-analyses have documented increased risk of mortality after an inpatient delirium admission 3 . To validate these findings in a larger population size while also accounting for more covariates than previously tested, including health status, we used our cohort of patients with delirium and their matched controls to conduct a longitudinal time-to-event analysis, where an event was defined as mortality or loss to follow-up. Kaplan-Meier survival curve visualization of the data showed different probabilities of survival rates between patients with an inpatient delirium admission (median 8.47 years; 95% CI, 7.8 to 9.35 in delirium group) versus control patients (Fig. 6 a). Cox proportional hazard ratio analysis unadjusted and adjusted for demographic characteristics (sex, age at admission, race, length of time in EHR, number of inpatient visits prior, total number of comorbidities) and visit features (type of visit, visit length, and length of follow-up time in EHR) revealed a significant increased risk of delirium in those with a delirium diagnosis compared to controls (HR 1.2; 95% CI, 1.16 to 1.29) (Fig. 6 b). A similar analysis for the UC-wide patient cohort also showed increased mortality in those with a delirium diagnosis compared to controls with a HR of 1.14 (95% CI, 1.1 to 1.18) (median 8.96 years; 95% CI, 8.16 to 10.1 in control group; median 3.83 years; 95% CI, 3.71 to 4.1 in delirium group) (Fig. 6 c,d). Discussion In the past few decades, clinical data from the EHR combined with integrative computational approaches have enabled the dissection of complex, heterogeneous diseases in an unbiased manner. This study aims to apply this strategy to better understand both risk factors and outcomes of patients who are diagnosed with delirium while in the hospital. We found that many well-known risk factors of delirium emerged through our analysis looking at comorbidities associated with an inpatient delirium diagnosis, including chronic neurological conditions such as dementia and epilepsy as well as acute conditions such as infections and metabolic disturbances. Our sex-stratified association studies revealed that, even within these large categories of diseases of the nervous system and infections, subtypes of conditions were sex-specific. We also validated several of these comorbidities associated with delirium, including anemia, bipolar disorder, and melanocytic nevus, through time-to-event analyses. Finally, we showed that a delirium diagnosis is independently associated with an increased risk of mortality, with UCSF and UC-wide data showing a median survival of ~ 8.5 and ~ 4 years, respectively. To our knowledge, this is the first study to conduct a deep analysis of delirium versus control patients using the EHR data with over 7,000 patients with delirium in the exploratory dataset and ~ 20,000 patients in the validation dataset. We aimed to find an appropriately matched control cohort by using propensity score matching and including covariates that assess patient health status as well as frequency of healthcare utilization. We also did not exclude patients based on prior cutoffs such as age thresholds or ICU versus non-ICU admissions, as done by previous studies. This strategy enabled us to include as many delirium cases as possible to find potential associations that may be subtle but meaningful. For instance, anemia before admission for delirium was enriched in delirium patients compared to controls in both UCSF and UC-wide datasets. This was also confirmed at the laboratory level, with significant mean hemoglobin difference in patients with and without delirium (11.6 versus 12.2 respectively). Our longitudinal analysis validated our comorbidity analysis and provided an understanding of the time between diagnosis of a risk factor to the development of delirium. In the UCSF data, for instance, roughly 3% of patients with anemia went on to develop delirium within 10 years of the diagnosis while less than 0.5% of patients without anemia developed delirium in this timeframe (hazard ratio of 9.4). One study in 700 patients specifically undergoing lumbar spinal fusion found that perioperative anemia is a risk factor for developing delirium post-operation 27 . Our study is the first to show that a diagnosis of anemia may be a more generalizable risk factor for developing inpatient delirium. Further studies need to be done to understand whether treating anemia more aggressively will reduce development of delirium. The large number of patients also enabled us to dissect the diagnostic associations with delirium in a sex-stratified manner. Our findings point to potentially new biological insights into the pathophysiology of delirium that are sex dependent. For instance, male patients had a significant association with Clostridioides difficile infections, an infection that often develops after antibiotic use. This could point to either the Clostridioides difficile toxin and the downstream consequences of the infection being associated with delirium or the antibiotic that was used that led to the infection being associated with delirium. Meanwhile, Klebsiella and E. coli infection were enriched in female patients with delirium, organisms that most commonly cause urinary tract infections, a source of infection more common in females compared to males. Dementia, one of the known risk factors of delirium, is also known to have sex-biased features, including in prevalence, clinical progression, and neuropathological findings 14 . Interestingly, our association studies found that subtypes of dementia were associated with delirium in a sex-dependent manner. Further studies will need to be done to understand whether this is due to the underlying sex differences in the prevalence of these dementia subtypes or sex differences in the pathophysiology of these dementias contributing to delirium. There are several limitations to our study that need further investigation. Given that illness-severity is a well-studied risk factor for delirium, we used several indirect metrics of health status, including number of inpatient admissions and number of comorbidities, to find an appropriately matched control cohort to our delirium cohort. In this way, we were able to compare groups with similar health statuses to extract more subtle risk factors that may be associated with delirium. Though we matched on the number of comorbidities to identify a control cohort with a similar health status, some comorbidities are more functionally debilitating and severe than others. Not all factors relating to delirium, or any condition of interest are captured or measured within the EHR as the data reflects only what is assessed or entered by clinicians. Given our control group had more pregnancy-related diagnoses, this suggests that though the control group may have a similar frequency of healthcare utilization, they are overall healthier than the matched delirium cohort. Future studies could use other health status metrics, such as the Charlson Comorbidity Index, which was not available in the datasets we used, or the American Society of Anesthesiologists (ASA) physical status classification score 28 . ASA was available in our UC-wide dataset and was significantly higher in patients with delirium compared to control patients ( Supplementary Table 4 ). The mean values, however, were 2.95 in control patients versus 3.05 in delirium patients, suggesting this difference is likely not clinically meaningful, given the scores are integer numbers. Defining an accurate delirium diagnosis in the EHR is challenging. Detection and clinical diagnosis of delirium itself is often challenging and likely underdiagnosed 2 , 29 . As a result, the published prevalence of delirium in older patients varies widely, with some estimates around 23% 30 while others as high as 88% in palliative care patients 31 . Prevalence of delirium in our study was lower than these estimates, likely due to our restricted definition of delirium only including one OMOP concept ID and including younger patients. This definition, however, likely missed patients that did not fit this exact delirium definition. Documentation of delirium is unfortunately inconsistent in the EHR even with a confirmed diagnosis, with one study finding that only 9 out of 25 cases coded by ICD-9 code 32 . RASS scores were available for some of our patients in the UC-wide dataset and showed scores were different between patients with delirium and controls, suggesting our inclusion criteria is likely consistent with clinical metrics of delirium. We, however, did not use RASS as a way to include or exclude patient as it is not a tool specific to delirium but rather an overall metric of sedation status. Future studies could include more patients by identifying delirium symptoms in clinical notes or including patients who have been screened for delirium using clinical tools such as the 4 A’s Test (4AT) or the confusion assessment method (CAM) 33 , 34 , which were not available in the datasets we used. In this study, the definition of sex in the EHR is likely a combination of sex assigned at birth, legal sex, and sex determined by the clinician. Documentation of gender identity remains a challenge in the EHR and only recently has garnered attention to provide more inclusive and affirmative health care for all patients 35 , 36 . Further studies will need to be done to understand whether the associations found in this study are different when taking into context biological sex versus patient-reported gender identity. Similarly, other social determinants of health were not taken into account in this study, such as the patient’s level of formal education, which can influence cognitive reserve, as well as family / caregiver dynamics. Overall, our study demonstrates the powerful application of the EHR to study heterogeneous disease processes and to better understand the risk factors and outcomes of disease at both a large patient population scale and a longitudinal time course. These results not only confirmed some of the known risk factors of delirium but also generated several new clinical hypotheses that will need to be further investigated. Applications of these analyses on other data types such as medication history may also expand the potential risk factors for delirium and increase the predictive power for the condition. These findings could also help develop future modeling studies for predicting which patients will develop delirium to focus prevention efforts towards these patients. Understanding the impact of sex differences in delirium remains understudied, and this study points to the importance of doing sex-stratified analyses and the potentially interesting pathophysiology of delirium that interacts with sex. Methods Resource Availability: Lead Contact: Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Marina Sirota ( [email protected] ). Materials availability: This study did not generate new unique reagents. Data and code availability: The UCSF EHR database is available to individuals affiliated with UCSF who can contact the UCSF’s Clinical and Translational Science Institute (CTSI) ( [email protected] ) or the UCSF’s Information Commons team for more information ( [email protected] ). The UC-wide EHR database is only available to UC researchers who have completed analyses in their respective UC first and have provided justification for scaling their analyses across UC health centers (more details at https://www.ucop.edu/uc-health/departments/center-for-data-driven-insights-and-innovations-cdi2.html or by contacting [email protected] . The code used for analysis will be made available on Github. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Experimental Model and Study Participation Details: All analysis of UCSF and UC-wide EHR data was performed under the approval of the Institutional Review Boards. All clinical data were de-identified and written informed consent was waived by the institutions. Patient cohorts were identified using the UCSF de-identified and UC-wide HIPAA-compliant limited data set OMOP EHR databases. The UCSF dataset included over five million patients from January 1, 1982 to February 20, 2023 while the UC-wide dataset included over seven million patients from January 1, 2012 to April 19, 2023 from 4 sites (UC San Diego, UC Los Angeles, UC Irvine, and UC Davis). Patients with inpatient delirium were identified using the OMOP concept ID 373995, SNOMED code 2776000 (corresponding to “Delirium”), filtered for first-time diagnosis of delirium during an inpatient stay (i.e., ‘visit of interest’). The inpatient control cohort with no delirium diagnosis was identified through propensity score (PS) matching (matchit R 37 ) by a generalized linear model at a 1:1 ratio using a nearest neighbor method and the following matching criteria: assigned sex, patient-reported race, estimated age at admission, years in EHR prior to visit, total number of comorbidities and inpatient visits prior to the visit of interest, stay length, stay type (ICU vs non-ICU), death during admission, and UC location (for UC-wide dataset). Propensity score matching was used similar to other previously published works using clinical datasets 17, 23, 38, 39 . For sex stratification, we utilized reported sex and excluded nonbinary and other categories due to low sample size, similar to a prior study 17 . See Table 1 for cohort demographic details. Method Details: Differential comorbidity analysis: Comorbidities were identified prior to the visit of interest with the earliest entry of every diagnosis. All patients and their comorbidities were visualized using Uniform Manifold Approximation and Projection (UMAP) using the R-implementation of the umap-learn package from Python 40 . UMAP is a form of dimensionality reduction to plot high-dimensional data into a lower-dimension. In this case, a table was created with all distinct comorbidities as columns and all patients as rows, with each table entry corresponding to 0 as the patient not having that diagnosis and 1 as the patient having that diagnosis. This table was then visualized as a UMAP with each point corresponding to one patient and encapsulating the similarity of that patient’s comorbidities to other patients based on distance. Correlations between variables (delirium status or sex) and UMAP coordinates were analyzed using Mann-Whitney U -tests. Differential comorbidity analysis between patients with delirium and controls was done using Fisher’s exact test and significance determined by Bonferroni-corrected threshold of p-value < 0.05. ICD10-CM blocks were also used to visualize differential comorbidity results. The OMOP concept table was used to map the SNOMED codes to ICD10 codes then mapped to categorical modules using the ICD10 codes. Overlaps between results of analyses using UCSF versus UC-wide datasets were done using hypergeometric test or Spearman correlation. A similar analysis was done for the sex-stratified analysis. Differential laboratory test results analysis: Laboratory test results were collected for tests done prior to the visit of interest, and the median values for all numerical lab tests was calculated. Lab tests that had more than 95% of patients missing results were excluded from the analysis, adapted from a prior study 17 . Lab value distributions were compared using Mann-Whitney U -tests across delirium status or sex. Longitudinal analyses: Time-to-event analysis was done by first identifying patients with a diagnosis of bipolar disorder (BD) and no prior delirium diagnosis. PS-matched control cohorts with no diagnosis of BD were identified using similar matching-criteria as described above. Time-to-event was calculated as time from first-time diagnosis of BD (or another non-BD diagnosis for the control patients) to either first inpatient delirium diagnosis, death, or loss to follow-up. Cox regression model was used to determine the hazard ratio, confidence intervals and significance (survival R 41 ). Time-to-event analysis for mortality after delirium was done using a similar strategy. Quantification and Statistical Analysis: All statistical analyses were performed using R Statistical Software version 4.3.2 (R Core Team 2023) and the specific statistical tests used for each experiment are outlined in detail above in each respective section as well as in the figure legends respectively. Declarations Acknowledgments We thank the UCSF Information Commons for their help navigating the UCSF EHR database, especially Evan Phelps, and the UC Health Data Warehouse team for their help with the UC-wide database, especially Robert Follett. This work was supported by the following grants: National Institutes of Health grant T32GM007618 (L.K., A.T.), National Institutes of Health grant R01AG060393 (M.S.), National Institutes of Health grant R01AG057683 (M.S.) Author contributions: Conceptualization: L.K., M.S. Methodology: L.K., S.W., A.T., Y.L., J.K., I.E.A., T.O., M.S. Investigation: L.K., M.S. Visualization: L.K. Funding acquisition: L.K., M.S. Project administration: L.K., M.S. Supervision: M.S. Writing – original draft: L.K., M.S. Writing – review & editing: L.K., S.W., A.T., Y.L., J.K., I.E.A., T.O., E.R., M.S. Competing interests: The authors declare no competing interests. References Association AP (2022) Diagnostic and statistical manual of mental disorders (5th ed., text rev.) Wilson JE et al (2020) Delirium. Nat Rev Dis Primers 6:90 Witlox J et al (2010) Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. JAMA 304:443–451 Fiest KM et al (2021) Long-Term Outcomes in ICU Patients with Delirium: A Population-based Cohort Study. Am J Respir Crit Care Med 204:412–420 Goldberg TE et al (2020) Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurol 77:1373–1381 Burry L et al (2018) Antipsychotics for treatment of delirium in hospitalised non-ICU patients. Cochrane Database Syst Rev 6:Cd005594 Burton JK et al (2021) Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients. Cochrane Database Syst Rev 7:Cd013307 Lindroth H et al (2018) Systematic review of prediction models for delirium in the older adult inpatient. BMJ Open 8:e019223 Baker M et al (2006) Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 442:916 Pendlebury ST et al (2016) Delirium risk stratification in consecutive unselected admissions to acute medicine: validation of externally derived risk scores. Age Ageing 45:60–65 Wong A et al (2018) Development and Validation of an Electronic Health Record-Based Machine Learning Model to Estimate Delirium Risk in Newly Hospitalized Patients Without Known Cognitive Impairment. JAMA Netw Open 1:e181018 Douglas VC et al (2013) The AWOL tool: derivation and validation of a delirium prediction rule. J Hosp Med 8:493–499 Solà-Miravete E et al (2018) Nursing assessment as an effective tool for the identification of delirium risk in older in-patients: A case-control study. J Clin Nurs 27:345–354 Lopez-Lee C, Kodama L, Gan L (2022) Sex Differences in Neurodegeneration: The Role of the Immune System in Humans. Biol Psychiatry 91:72–80 Trzepacz PT et al (2018) Delirium Phenotype by Age and Sex in a Pooled Data Set of Adult Patients. J Neuropsychiatry Clin Neurosci 30:294–301 Wiredu K et al (2023) Sex Differences in the Incidence of Postoperative Delirium After Cardiac Surgery: A Pooled Analyses of Clinical Trials. Anesthesiology Tang AS et al (2022) Deep phenotyping of Alzheimer's disease leveraging electronic medical records identifies sex-specific clinical associations. Nat Commun 13:675 Li L et al (2015) Identification of type 2 diabetes subgroups through topological analysis of patient similarity. Sci Transl Med 7:311ra174 Abraham A et al (2022) Dense phenotyping from electronic health records enables machine learning-based prediction of preterm birth. BMC Med 20:333 Månsson R, Joffe MM, Sun W, Hennessy S (2007) On the estimation and use of propensity scores in case-control and case-cohort studies. Am J Epidemiol 166:332–339 Rose S, Laan MJ (2009) Why match? Investigating matched case-control study designs with causal effect estimation. Int J Biostat 5:1 Weberpals J et al (2021) Deep Learning-based Propensity Scores for Confounding Control in Comparative Effectiveness Research: A Large-scale, Real-world Data Study. Epidemiology 32:378–388 Woldemariam SR, Tang AS, Oskotsky TT, Yaffe K, Sirota M (2023) Similarities and differences in Alzheimer's dementia comorbidities in racialized populations identified from electronic medical records. Commun Med (Lond) 3:50 Huguet N et al (2020) Using Electronic Health Records in Longitudinal Studies: Estimating Patient Attrition. Med Care 58(Suppl 1):S46–s52 Ely EW et al (2003) Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 289:2983–2991 Han JH et al (2015) The Diagnostic Performance of the Richmond Agitation Sedation Scale for Detecting Delirium in Older Emergency Department Patients. Acad Emerg Med 22:878–882 Gold C et al (2022) Risk factors for delirium in elderly patients after lumbar spinal fusion. Clin Neurol Neurosurg 219:107318 Hurwitz EE et al (2017) Adding Examples to the ASA-Physical Status Classification Improves Correct Assignment to Patients. Anesthesiology 126:614–622 Delirium is prevalent in (2019) older hospital inpatients and associated with adverse outcomes: results of a prospective multi-centre study on World Delirium Awareness Day. BMC Med 17:229 Gibb K et al (2020) The consistent burden in published estimates of delirium occurrence in medical inpatients over four decades: a systematic review and meta-analysis study. Age Ageing 49:352–360 Hosie A, Davidson PM, Agar M, Sanderson CR, Phillips J (2013) Delirium prevalence, incidence, and implications for screening in specialist palliative care inpatient settings: a systematic review. Palliat Med 27:486–498 Hope C et al (2014) Documentation of delirium in the VA electronic health record. BMC Res Notes 7:208 Bellelli G et al (2014) Validation of the 4AT, a new instrument for rapid delirium screening: a study in 234 hospitalised older people. Age Ageing 43:496–502 Inouye SK et al (1990) Clarifying confusion: the confusion assessment method. A new method for detection of delirium. Ann Intern Med 113:941–948 Patel K, Lyon ME, Luu HS (2021) Providing Inclusive Care for Transgender Patients: Capturing Sex and Gender in the Electronic Medical Record. J Appl Lab Med 6:210–218 Lau F, Antonio M, Davison K, Queen R, Devor A (2020) A rapid review of gender, sex, and sexual orientation documentation in electronic health records. J Am Med Inf Assoc 27:1774–1783 Ho DI, King K, Stuart G (2011) MatchIt: Nonparametric Preprocessing for Parametric Causal Inference. J Stat Softw 42:1–28 Marić I et al (2021) Decreased Mortality Rate Among COVID-19 Patients Prescribed Statins: Data From Electronic Health Records in the US. Front Med (Lausanne) 8:639804 Oskotsky T et al (2021) Mortality Risk Among Patients With COVID-19 Prescribed Selective Serotonin Reuptake Inhibitor Antidepressants. JAMA Netw Open 4:e2133090 McInnes LH, Melville J (2020) J. UMAP: uniform manifold approximation and projection for dimension reduction. arXiv:1802.03426 Therneau TMG (2000) P. M. Modeling Survival Data: Extending the Cox Model New York Additional Declarations There is NO Competing Interest. Supplementary Files Combinedsupplementaryfiguresfinal.pdf SupplementaryTables.xlsx Supplementary Table 1 (Related to Fig. 2): Differential comorbidities between delirium and control patients in UCSF data Supplementary Table 2 (Related to Fig. 2): Differential comorbidities between delirium and control patients in UC-wide data Supplementary Table 3 (Related to Fig. 2): Differential comorbidities between delirium and control patients significant in both datasets Supplementary Table 4 (Related to Fig. 4): Sex-stratified differential comorbidities between delirium and control patients significant in UCSF data Supplementary Table 5 (Related to Fig. 4): Sex-stratified differential comorbidities between delirium and control patients significant in UC-wide data Cite Share Download PDF Status: Published Journal Publication published 22 Jul, 2025 Read the published version in Communications Medicine → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4765249","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":330103748,"identity":"fd3045ef-07f1-4c23-9bfd-81c440605da2","order_by":0,"name":"Marina Sirota","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+klEQVRIiWNgGAWjYBACgwMMDEDEYM8G5lYAMTOQz4NHiyVUS2IbAwNjA8MZIJMtAb8W+wMQGqgMqIWxjQgtZsfPPjxcUMOQwCd9+PiDn/MOy8u3MTA+eNuGR8uZdIPDM44B/cKXltjYu+2w4YZjDMyGc/FpOZDGcJiHjYGxjYfHsIF32+EEA/kGNmlePFoMzj8DavkH0dL4d87hBKDD2H/j1XIDaAtQAVhLM2/D4QSGYwxszPi1AG3h7ZNIbONhS5wtcywd6BfGZsk55/A5LI35M883G3v5HuYDH9/UWANDjPnghzdluLVAgQQyBxilo2AUjIJRMAooAwDT5FA2oqxLPgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-7246-6083","institution":"University of California, San Francisco","correspondingAuthor":true,"prefix":"","firstName":"Marina","middleName":"","lastName":"Sirota","suffix":""},{"id":330103751,"identity":"180bdbf2-6608-43cf-81bb-dfc6cda4e27f","order_by":1,"name":"Lay Kodama","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Lay","middleName":"","lastName":"Kodama","suffix":""},{"id":330103752,"identity":"7bb7168b-289f-4b96-ad97-7c92cf838ee6","order_by":2,"name":"Sarah Woldemariam","email":"","orcid":"https://orcid.org/0000-0003-1873-8575","institution":"University of California San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Woldemariam","suffix":""},{"id":330103753,"identity":"201ab815-3814-4359-afb4-c97e1e0a17af","order_by":3,"name":"Alice Tang","email":"","orcid":"https://orcid.org/0000-0003-4745-0714","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Alice","middleName":"","lastName":"Tang","suffix":""},{"id":330103754,"identity":"4f612d51-f684-4361-8851-efe63937ff01","order_by":4,"name":"Yaqiao Li","email":"","orcid":"","institution":"Gladstone Institutes","correspondingAuthor":false,"prefix":"","firstName":"Yaqiao","middleName":"","lastName":"Li","suffix":""},{"id":330103756,"identity":"58afa88e-aa03-4ded-9af8-1c7927f74a90","order_by":5,"name":"John Kornak","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Kornak","suffix":""},{"id":330103758,"identity":"cd047409-f1e2-4c88-a7d6-6a14bc39164a","order_by":6,"name":"Isabel (E) Allen","email":"","orcid":"","institution":"UCSF","correspondingAuthor":false,"prefix":"","firstName":"Isabel","middleName":"(E)","lastName":"Allen","suffix":""},{"id":330103759,"identity":"d941058f-fa50-44f6-964a-8e9bb5d35b88","order_by":7,"name":"Eva Raphael","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Eva","middleName":"","lastName":"Raphael","suffix":""},{"id":330103762,"identity":"0b5cc651-a270-4f38-9986-f0d0b98aa86d","order_by":8,"name":"Tomiko Oskotsky","email":"","orcid":"https://orcid.org/0000-0001-7393-5120","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Tomiko","middleName":"","lastName":"Oskotsky","suffix":""}],"badges":[],"createdAt":"2024-07-18 23:05:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4765249/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4765249/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43856-025-00986-5","type":"published","date":"2025-07-22T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":60911143,"identity":"9fd3ed4e-1373-467a-97a0-6aa11e330f58","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":255362,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic of study. \u003c/strong\u003ePatient cohorts were selected from the OMOP electronic health record databases from UC San Francisco (UCSF) and UC-Wide databases (4 sites: UC Davis, UC San Diego, UC Los Angeles, UC Irvine). All patients and their first inpatient diagnosis of delirium were selected and a control cohort with no delirium diagnosis was selected using propensity score matching on the features listed in the formula. Association and longitudinal analyses were done using these cohorts. Association studies were done with prior diagnoses and prior laboratory results, with and without sex-stratification. Longitudinal analyses included time-to-delirium diagnosis after first-time diagnosis of selected potential diagnostic risk factor and time-to-mortality outcome after an inpatient delirium diagnosis. See also Table 1.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/65eaeb872a94033a9b2ce133.png"},{"id":60913289,"identity":"340d7f36-5832-454f-88eb-0f596f91e0cc","added_by":"auto","created_at":"2024-07-23 13:09:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":302192,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDiagnostic associations with delirium show enrichment of known risk factors of delirium. (a)\u003c/strong\u003e UMAP representation of all first-time, non-delirium diagnoses prior to the inpatient visit of interest. Each dot represents a patient (salmon = patient with delirium, grey = control patient). Violin plots showing distribution of patients across UMAP component 1 (left) and 2 (right). P-values determined by two-sided Mann–Whitney U-tests. **** = p-value \u0026lt; 2.2e-16; ** = p-value 0.0086. \u003cstrong\u003e(b) \u003c/strong\u003eVolcano plot of differential comorbidities, with diagnoses enriched in controls in black (108 diagnoses) and in delirium patients in salmon (101 diagnoses) and non-significant diagnoses in grey. Significance determined by two-sided Fisher’s exact test with Bonferroni-corrected p-value \u0026lt; 0.05 (at dotted horizontal line). OR = odds ratio. Most significant diagnoses highlighted by name. \u003cstrong\u003e(c)\u003c/strong\u003e ICD10-diagnostic block representation of significant differential comorbidities identified in (b) for patients with delirium. \u003cstrong\u003e(d)\u003c/strong\u003eTable showing number of diagnoses overlapping between UCSF and UC-Wide datasets in each ICD10 block for each patient group. Entries with at least 3 or more diagnoses in one patient group compared to the other are colored. Hypergeometric test used to evaluate overlap between the two datasets (p-value 1.2 e-94). \u003cstrong\u003e(e)\u003c/strong\u003e Log-log plot comparing differential diagnoses between UCSF and UC-Wide databases. Only plotting diagnoses significant in either (grey) or both (salmon) databases. Spearman correlation 𝜌 = 0.94 when looking at points significant in both datasets. \u003cstrong\u003e(f) \u003c/strong\u003eZoomed in plot of the yellow-highlighted portion of plot in (e). Significant comorbidities found in both databases with diagnoses with the largest odds ratios highlighted. See also Extended Data Fig. 3.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/e9e22306d27953e659a49538.png"},{"id":60911142,"identity":"58e539a5-e076-4a6e-9ed1-75f81e3809b8","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":135504,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of differential laboratory tests between delirium and control patients. (a) \u003c/strong\u003eLog-log plot comparing all significantly differential laboratory results between delirium versus control patients in UCSF (y-axis) and UC-Wide (x-axis) databases. Axes reflect log base 2 of the mean lab values for delirium versus control patients.\u003cstrong\u003e (b,c)\u003c/strong\u003e. Violin plots showing distribution of patients and their mean hemoglobin and hematocrit values for UCSF (b) and UC-wide (c) datasets. Black point denotes mean of population. Two-sided Mann-Whitney U-test, **** = p-value \u0026lt; 2.2e-16.\u003cstrong\u003e \u003c/strong\u003eSee also Extended Data Fig. 4.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/4f3851fde37df1900abb0179.png"},{"id":60912352,"identity":"9ff3800f-d7ec-4986-92ea-96456af01c04","added_by":"auto","created_at":"2024-07-23 13:01:46","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":168439,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of sex-specific diagnostic associations with delirium, including dementia subtypes and infections. \u003c/strong\u003eLog-log plots comparing differential diagnostic associations with delirium in UCSF (x-axis) versus UC-wide (y-axis) databases. Plots split by diagnoses significant only in females (a), only in males (b), or in both (c). Diagnoses with largest OR values highlighted. Spearman correlation 𝜌 = 0.9 (a), 0.35 (b), 0.55 (c). See also Extended Data Fig. 4 and Extended Data Fig. 5.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/5150c62e5bd7d8cd4b8ce565.png"},{"id":60911147,"identity":"f292b72d-85c7-408e-80d2-1ef149dff062","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":152204,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA prior diagnosis of anemia or bipolar disorder leads to increased risk of developing delirium. \u003c/strong\u003eKaplan-Meier curve showing time-to-event where event is defined as delirium, death, or loss to follow-up since the first-time diagnosis of anemia (a), bipolar disorder (c), and melanocytic nevus (e) in UCSF patients. Cox proportional hazard ratio analysis done for anemia (b), bipolar disorder (d), and melanocytic nevus (f) with unadjusted and adjusted analyses (adjusted for sex, age at admission, race, length of time in EHR, number of inpatient visits prior, total number of comorbidities, and length of follow-up time in EHR) in UCSF (top) and UC-wide (bottom) patients. Confidence intervals representing 95% CI. See also Extended Data Fig. 6 and Extended Data Fig. 7.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/6ad31b686cb478930486b73e.png"},{"id":60911148,"identity":"864c3594-3cb3-4122-a343-b08d9efb7f2a","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":103643,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIncreased mortality outcome after an inpatient delirium diagnosis. \u003c/strong\u003eKaplan-Meier survival curve showing time-to-death after inpatient admission of interest for control (grey) and delirium (salmon) patients in UCSF (a) and UC-Wide (c) data. Follow up time periods differ between the two datasets given the discrepancy in the number of years captured by each dataset. Cox proportional hazard ratio analysis done for in UCSF (b) and UC-wide (d) cohort with unadjusted and adjusted analyses (adjusted for sex, age at admission, race, length of time in EHR, number of inpatient visits prior, total number of comorbidities, type of visit, visit length, and length of follow-up time in EHR).\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/8f61dbdba86e5c3e6082f9b2.png"},{"id":87371192,"identity":"ae91b431-6e63-413b-9e47-74fd38303196","added_by":"auto","created_at":"2025-07-23 07:12:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2630621,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/d7e3a011-51cf-463d-9263-fbf8d018125a.pdf"},{"id":60911150,"identity":"1d0426bf-78fe-483e-bf42-5dce141606ba","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7452523,"visible":true,"origin":"","legend":"","description":"","filename":"Combinedsupplementaryfiguresfinal.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/b359a43b327ab48df5c7f5fc.pdf"},{"id":60911149,"identity":"312c98a8-8e09-409b-a65a-7a0482af7e7d","added_by":"auto","created_at":"2024-07-23 12:53:46","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":926307,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Table 1 (Related to Fig. 2): Differential comorbidities between delirium and control patients in UCSF data\u003c/p\u003e\n\u003cp\u003eSupplementary Table 2 (Related to Fig. 2): Differential comorbidities between delirium and control patients in UC-wide data\u003c/p\u003e\n\u003cp\u003eSupplementary Table 3 (Related to Fig. 2): Differential comorbidities between delirium and control patients significant in both datasets\u003c/p\u003e\n\u003cp\u003eSupplementary Table 4 (Related to Fig. 4): Sex-stratified differential comorbidities between delirium and control patients significant in UCSF data\u003c/p\u003e\n\u003cp\u003eSupplementary Table 5 (Related to Fig. 4): Sex-stratified differential comorbidities between delirium and control patients significant in UC-wide data\u003c/p\u003e","description":"","filename":"SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4765249/v1/6a980f05a120de45ffe68d4f.xlsx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Sex-stratified analyses of comorbidities associated with an inpatient delirium diagnosis using real world data","fulltext":[{"header":"Main","content":"\u003cp\u003eDelirium is a clinical diagnosis defined as fluctuating disturbances in attention, awareness, and cognition that develops over a short period of time and is highly prevalent among older inpatient populations, with an estimated prevalence of 23% in hospitalized older adults \u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Both long- and short-term outcomes of delirium are detrimental to patients, distressing to caregivers, and a burden on the healthcare system \u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Notably, patients who are diagnosed with delirium have an increased risk of long-term mortality. It is unclear, however, whether delirium is an independent risk factor for mortality versus other confounding factors are at play such as overall baseline disease burden \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. Medications for treating delirium are largely for symptomatic management and have not been shown to have clinical benefits and in fact may lead to worse outcomes, especially for older patients \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Prevention is the most effective strategy and often involve non-pharmacological interventions, such as reorientation and cognitive stimulation \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, making it imperative to predict which patients may develop delirium using predictive tools to focus prevention efforts for high-risk patients.\u003c/p\u003e \u003cp\u003eThough prevalent, delirium remains challenging to study and predict given its heterogenous clinical nature and diverse risk factors \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Current prediction tools have variable predictive capabilities, with one systematic review finding an area under the receiver operating curve range from 0.52 to 0.94 \u003csup\u003e8\u003c/sup\u003e. These models have other limitations including validation in small sample sizes or data collected from one institution that may not be generalizable \u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These models incorporate well-studied risk factors such as existing cognitive impairment and severity of chronic illnesses, but less apparent risk factors also need to be identified to improve these predictive tools \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eImportantly, sex is a major risk modifier for many neurological diseases, with some evidence of sex-differences in delirium \u003csup\u003e\u003cspan additionalcitationids=\"CR15\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. For instance, one study found hypoactive delirium to be more common in female compared to male patients \u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Though clinical sex differences have been well-documented in dementia and cognitive impairment -- major risk factors for delirium -- whether sex differences exist in delirium remains largely unstudied due to the lack of sex-stratified studies.\u003c/p\u003e \u003cp\u003eOne strategy to overcome these limitations is to employ large-scale, comprehensive real-world data from electronic health records (EHR) combined with robust computational approaches. Data from the EHR have been used to better understand and phenotype complex diseases such as Alzheimer\u0026rsquo;s disease, type 2 diabetes, and preterm birth \u003csup\u003e\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Such deep phenotyping studies can shed insights into clinical risk factors and subtypes of disease as well as point to potential new biological pathways that may be involved in disease.\u003c/p\u003e \u003cp\u003eIn this study, we leveraged EHRs from two databases across California to identify differential comorbidities and laboratory test results prior to a patient\u0026rsquo;s first inpatient admission for delirium, which could serve as potential factors in identifying patients at risk for the condition. Importantly, we also conducted a sex-stratified analysis to identify differences by documented sex in the association between comorbidities or laboratory findings with a delirium diagnosis. We further analyzed specific comorbidities using a longitudinal approach to predict the length of time from the diagnosis of a comorbidity and admission for delirium, as an alternative, complementary approach to validate the association study and also provide granularity into the time course from comorbidity diagnosis to delirium diagnosis. Lastly, we conducted a longitudinal analysis to understand the relationship between first inpatient delirium diagnosis and mortality outcomes.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eIdentification of patients with a diagnosis of delirium and their matched controls\u003c/h2\u003e \u003cp\u003eWe identified 7,492 patients with an inpatient delirium diagnosis and 7,492 propensity-score (PS)-matched control patients within the University of California, San Francisco (UCSF) de-identified electronic health record (EHR) database (~\u0026thinsp;5\u0026nbsp;million patients total). Propensity score matching was used to create a matched case-control study as individual covariates may not be well-matched but the \u0026ldquo;propensity\u0026rdquo; to be either a case or control is well-matched, as illustrated in prior publications justifying this method \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Control patients were matched on the following demographic and inpatient visit features: age at admission, patient-identified race, sex, death during admission, length of the inpatient visit of interest in days, years of available EHR data, total number of inpatient visits prior to the visit of interest, total number of comorbidities prior to the visit of interest, and whether the visit of interest was in the ICU setting. The \u0026ldquo;visit of interest\u0026rdquo; corresponded to the visit where an inpatient delirium diagnosis was made for the delirium group or a randomly selected inpatient visit for the control group (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Similarly, a separate cohort of 19,417 patients with an inpatient delirium diagnosis and 19,417 PS-matched control cohort were identified from the UC-Wide EHR database (~\u0026thinsp;8.6\u0026nbsp;million patients total, data from UC Davis, UC Los Angeles, UC Irvine, UC San Diego) with an additional matching criterion that included UC location (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These covariates were chosen based on prior EHR-based studies using PS-matched cohort selections with inclusion of additional covariates to control for health status and healthcare utilization frequency using indirect measures such as the total number of inpatient visits and comorbidities prior to the visit of interest and years of available EHR data \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan additionalcitationids=\"CR23\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Because patients who develop delirium tend to be at baseline more ill than those who do not develop delirium, we wanted to find a comparable control group that had similar baseline disease burden and identify what specific comorbidities are differentially enriched between patients who are similarly \"ill\", but one group develops delirium and the other group does not.\u003c/p\u003e \u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e\u003cstrong\u003eDemographic information of matched cohorts.\u003c/strong\u003e Table of demographic information for patient cohorts identified in UCSF (left) and UC-wide (right) data. Chi-squared test used for categorical measures. Student\u0026rsquo;s t-test used for continuous measures. SMD = standardized mean difference. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"780\" height=\"262\"\u003e\u003c/p\u003e\u003cp\u003ePost-matching analysis showed adequate matching of covariates with similar PS distributions between the groups (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,c) and absolute standardized mean differences less than 0.1, except for inpatient stay length which had a wide distribution in both databases (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb,d). Detailed demographic and inpatient visit features for the cohorts generated are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. We defined inpatient delirium broadly using the Observational Medical Outcomes Partnership (OMOP) concept ID 373995 (corresponding to \u0026ldquo;Delirium\u0026rdquo;) and excluded diagnoses that had specific causes of delirium in the diagnosis name (such as alcohol-induced delirium or other substance-induced delirium). These exclusion criteria were used to focus on cases where a delirium diagnosis was made but the cause of the delirium was not readily known. Using this definition, we were able to capture 84% and 89% of all delirium-related visits in the UCSF and UC-wide databases, respectively, with delirium prevalence within the wide range of published estimates for patients 65 years and older (9.5% mean, 8.9% SD in UCSF data; 3.5% mean, 1.9% SD in UC-wide data) (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea,b).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe used the Richmond Agitation Sedation Scale (RASS) rating to corroborate the delirium diagnosis in the UC-wide data. RASS is a 10-point rating scale ranging from \u0026minus;\u0026thinsp;5 to 5, with negative numbers corresponding to the level of sedation and positive numbers corresponding to the level of agitation. Though RASS is a more general measure of a patient\u0026rsquo;s level of sedation, validated mostly in the ICU setting, it has also been implemented more widely in the inpatient setting for detection of delirium \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. Less than half of patients had documented RASS ratings during their visit of interest. Of the available ratings, patients with delirium had a statistically significant increase in the mean RASS rating compared to controls (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), suggesting patients with a delirium diagnosis are overall more agitated during the visit and also have greater variability in their RASS ratings (data not shown), suggesting a waxing and waning nature to their clinical status.\u003c/p\u003e \u003cp\u003e \u003cb\u003ePatients with delirium are more likely to be diagnosed with diseases of the nervous system, mental health, metabolic disorders, and infections compared to control patients\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo understand potential risk factors associated with an inpatient delirium diagnosis, we first collected all first-time diagnoses made during visits prior to the inpatient admission of interest. Low-dimensional Uniform Manifold Approximation and Projection (UMAP) representation of all non-delirium diagnoses (19,590 features, SNOMED concept IDs) shows a statistically significant separation of patients with a delirium diagnosis versus matched controls by two-sided Mann-Whitney U test (UMAP 1, p-value\u0026thinsp;\u0026lt;\u0026thinsp;2.2 e-16; UMAP 2 p-value 0.0086; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eDifferential association analyses of these comorbidities using Fisher\u0026rsquo;s exact test showed enrichment of distinct comorbidities for patients with delirium compared to control patients, with 101 diagnoses significantly enriched in patients with delirium versus 108 in controls, out of 19,583 diagnoses tested (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Control patients had enrichment of diagnoses largely related to pregnancy and other health statuses as well as age-related musculoskeletal diagnoses such as pain in joints and osteoarthritis and skin findings such as melanocytic nevus (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb,c, \u003cb\u003eand Supplementary Table\u0026nbsp;1\u003c/b\u003e). Meanwhile, patients with delirium had enrichment of diagnoses related to diseases of the nervous system, including epilepsy and seizures, mental health and behavioral disorders, and acute diagnoses such as metabolic diseases and infections (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb,c, \u003cb\u003eand Supplementary Table\u0026nbsp;1\u003c/b\u003e). Similar diagnostic associations were seen in the UC-wide cohort based on a hypergeometric test (p-value 1.2 e-94; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed-f, \u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, \u003cb\u003eSupplementary Table\u0026nbsp;2, and Supplementary Table\u0026nbsp;3\u003c/b\u003e). Categorizations of these overlapping diagnoses between the two databases by ICD10 diagnostic blocks showed enrichment of diagnoses related to certain disease categories in patients with delirium versus control patients, including blood-related diseases such as anemia, diseases of the genitourinary system such as urinary tract infections, diseases of the nervous system such as epilepsy, metabolic diseases such as hyponatremia, and mental health-related disorders such as bipolar disorder (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed \u003cb\u003eand Supplementary Table\u0026nbsp;3\u003c/b\u003e). These findings are largely consistent with previously-identified risk factors for delirium \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eDifferential laboratory findings corroborate differential associations between comorbidities and delirium\u003c/h2\u003e \u003cp\u003eWe also conducted enrichment analysis for mean laboratory values and vital signs collected before the inpatient admission of interest. Patients with a delirium diagnosis had significantly higher mean values of certain liver function tests, such as alkaline phosphatase and aspartate transferase, compared to controls in both UCSF and UC-wide datasets, suggesting potential liver dysfunction in these patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Elevated mean vital signs included heart rate and respiratory rate \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Meanwhile, glomerular filtration rate and urine creatinine levels were decreased in patients with delirium compared to controls, consistent with kidney dysfunction (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea\u003cb\u003e).\u003c/b\u003e Hemoglobin, hematocrit, and erythrocyte counts were also significantly decreased in patients with delirium compared to controls, consistent with the association with an anemia diagnosis in these patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The UC-wide dataset also captured certain clinical test scores, including results from the Patient Health Questionnaire (PHQ)-2 and PHQ-9, consistent with the associations with depression in these patients (\u003cb\u003eSupplementary Table\u0026nbsp;4\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eSex-stratified analysis shows certain infections and dementia subtypes are sex-specific risk factors for delirium\u003c/h2\u003e \u003cp\u003eTo understand whether any of the comorbidities associated with delirium that we identified are sex-specific, we conducted a sex-stratified association analysis using the same cohort identified above (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e \u003cb\u003eand Extended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). We identified several diagnoses that were significantly associated with delirium in only females, only males, or in both female and male patients with delirium compared to controls in both UCSF and UC-wide datasets (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec-e \u003cb\u003eand Extended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec-e) as well as sex-specific laboratory results (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef \u003cb\u003eand Extended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDiagnoses common to both male and female patients with delirium in both datasets were largely similar to the non-stratified analysis, including symptoms of delirium such as altered mental status, restlessness and agitation, and hallucinations, as well as known organic causes of delirium and altered mental status such as seizures, cognitive disorders and other mental disorders (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). Interestingly, female and male patients diagnosed with delirium had sex-specific associations with distinct infections and diseases of the nervous system that were statistically significant in both datasets. For instance, female patients with delirium had associations with encapsulated bacterial infections due to Streptococcal bacteria (OR UCSF 2.38; UC-wide 1.61), \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e (OR UCSF 2.95, UC-wide 2.14), \u003cem\u003eEscherichia coli\u003c/em\u003e (OR UCSF 1.45, UC-wide 1.81), and enterococcal bacteria (OR UCSF 1.53, UC-wide 1.79), while males had associations with \u003cem\u003eClostridioides difficile\u003c/em\u003e infections (OR UCSF 3.89, UC-wide 1.55) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,b). Sex-specific associations with subtypes of dementia were also seen, where females had significant associations with Alzheimer\u0026rsquo;s disease (OR UCSF 2.06, UC-wide 3.55) and vascular dementias with behavioral disturbances (OR UCSF 1.89, UC-wide 3.29) while males had associations with Diffuse Lewy Body disease (OR UCSF 1.51, UC-wide 4.22) and diagnoses related to symptoms of the disease such as visual hallucinations, falls, difficulty walking / muscle weakness, dysphagia, insomnia, and constipation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,b, \u003cb\u003eSupplementary Table\u0026nbsp;5, and Supplementary Table\u0026nbsp;6\u003c/b\u003e). These sex-specific associations could be because these dementia subtypes are known to have sex-differences in their prevalence already \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e or could point to sex-specific ways in which delirium manifest in different patient populations.\u003c/p\u003e \u003cp\u003eThough there were several sex-specific laboratory results found in the UCSF and UC-wide datasets, none were common between the two datasets. Laboratory findings associated with delirium in both male and female patients as well as between both datasets were similar to the results of the non-sex-stratified analysis, including decreased hemoglobin and hematocrit consistent with an anemia diagnosis, decreased GFR consistent with kidney dysfunction, elevated liver function tests such as alkaline phosphatase, and elevation of heart rate (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea, \u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef, \u003cb\u003eand Extended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef).\u003c/p\u003e \u003cp\u003e \u003cb\u003eLongitudinal analysis from diagnosis of a potential risk factor of delirium to an inpatient delirium diagnosis validates comorbidity association study\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo further understand several of the comorbidities associated with an inpatient delirium diagnosis identified above, we carried out a longitudinal time-to-event analysis for anemia and bipolar disorder. We chose these diagnoses for analysis given the higher association of blood-related disorders and mental and behavioral disorders in patients with delirium found through our association study (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Patients with a first-time diagnosis of anemia and no prior diagnosis of delirium and a matched control group with no diagnosis of anemia and no prior diagnosis of delirium were identified. Control patients were matched on the following parameters: age at the visit of interest, patient-identified race, documented sex, years of available EHR data, total number of inpatient visits prior to the visit of interest, and total number of comorbidities prior to the visit of interest (\u003cb\u003eExtended Data\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Events were defined as admission for delirium, death, or loss to follow-up. Kaplan-Meier curve visualization of the data showed stratification by anemia diagnosis status, where those with anemia have increased probability of developing first-time inpatient delirium diagnosis (UCSF 3.4%, UC-wide 1.3% of anemia patients) than those without any anemia diagnosis (UCSF 0.3%, UC-wide 0.3% of control patients) over the course of ~\u0026thinsp;30 years in the UCSF data and ~\u0026thinsp;11 years in the UC-wide data (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea \u003cb\u003eand Extended Data Fig.\u0026nbsp;7a\u003c/b\u003e). Cox proportional hazard ratio analysis unadjusted and adjusted for demographics and visit features revealed a significant increased risk of delirium in those with an anemia diagnosis compared to controls (UCSF HR 9.4; 95% CI, 8.1 to 11; UC-wide HR 4.4; 95% CI, 4.1 to 4.7) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). A similar analysis was done for patients with and without Bipolar I Disorder (BD1) in UCSF patients or bipolar disorder (unspecified type) in UC-wide patients. We found that a diagnosis of BD also increased risk of developing first-time inpatient delirium diagnosis (UCSF 1.9%, UC-wide 0.7% of BD patients) than those without a BD diagnosis (UCSF 0.01%, UC-wide 0.01% of control patients) with a HR of 27 (95% CI, 9.9 to 74.4) for UCSF patients and HR of 7.8 (95% CI, 6.0 to 10.0) for UC-wide patients over the course of ~\u0026thinsp;20 years in the UCSF data and ~\u0026thinsp;10 years in the UC-wide data (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec,d, \u003cb\u003eand Extended Data Fig.\u0026nbsp;7b\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eOur association studies also found that certain diagnoses are more enriched in control patients compared to patients with delirium, including diagnoses under the ICD10-CM neoplasm category such as melanocytic nevus (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed \u003cb\u003eand Supplementary Table\u0026nbsp;3\u003c/b\u003e). We did a similar time-to-event analysis of patients with and without this diagnosis and found that, indeed, a prior diagnosis of melanocytic nevus modestly decreased the risk of developing delirium at a HR of 0.3 (95% CI, 0.2 to 0.5) for UCSF patients and HR of 0.76 (95% CI, 0.67 to 0.86) for UC-wide patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee,f, \u003cb\u003eand Extended Data Fig.\u0026nbsp;7c\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eA single inpatient delirium admission is associated with increased mortality\u003c/h2\u003e \u003cp\u003ePrevious meta-analyses have documented increased risk of mortality after an inpatient delirium admission \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. To validate these findings in a larger population size while also accounting for more covariates than previously tested, including health status, we used our cohort of patients with delirium and their matched controls to conduct a longitudinal time-to-event analysis, where an event was defined as mortality or loss to follow-up. Kaplan-Meier survival curve visualization of the data showed different probabilities of survival rates between patients with an inpatient delirium admission (median 8.47 years; 95% CI, 7.8 to 9.35 in delirium group) versus control patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). Cox proportional hazard ratio analysis unadjusted and adjusted for demographic characteristics (sex, age at admission, race, length of time in EHR, number of inpatient visits prior, total number of comorbidities) and visit features (type of visit, visit length, and length of follow-up time in EHR) revealed a significant increased risk of delirium in those with a delirium diagnosis compared to controls (HR 1.2; 95% CI, 1.16 to 1.29) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). A similar analysis for the UC-wide patient cohort also showed increased mortality in those with a delirium diagnosis compared to controls with a HR of 1.14 (95% CI, 1.1 to 1.18) (median 8.96 years; 95% CI, 8.16 to 10.1 in control group; median 3.83 years; 95% CI, 3.71 to 4.1 in delirium group) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec,d).\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn the past few decades, clinical data from the EHR combined with integrative computational approaches have enabled the dissection of complex, heterogeneous diseases in an unbiased manner. This study aims to apply this strategy to better understand both risk factors and outcomes of patients who are diagnosed with delirium while in the hospital. We found that many well-known risk factors of delirium emerged through our analysis looking at comorbidities associated with an inpatient delirium diagnosis, including chronic neurological conditions such as dementia and epilepsy as well as acute conditions such as infections and metabolic disturbances. Our sex-stratified association studies revealed that, even within these large categories of diseases of the nervous system and infections, subtypes of conditions were sex-specific. We also validated several of these comorbidities associated with delirium, including anemia, bipolar disorder, and melanocytic nevus, through time-to-event analyses. Finally, we showed that a delirium diagnosis is independently associated with an increased risk of mortality, with UCSF and UC-wide data showing a median survival of ~\u0026thinsp;8.5 and ~\u0026thinsp;4 years, respectively.\u003c/p\u003e \u003cp\u003eTo our knowledge, this is the first study to conduct a deep analysis of delirium versus control patients using the EHR data with over 7,000 patients with delirium in the exploratory dataset and ~\u0026thinsp;20,000 patients in the validation dataset. We aimed to find an appropriately matched control cohort by using propensity score matching and including covariates that assess patient health status as well as frequency of healthcare utilization. We also did not exclude patients based on prior cutoffs such as age thresholds or ICU versus non-ICU admissions, as done by previous studies. This strategy enabled us to include as many delirium cases as possible to find potential associations that may be subtle but meaningful. For instance, anemia before admission for delirium was enriched in delirium patients compared to controls in both UCSF and UC-wide datasets. This was also confirmed at the laboratory level, with significant mean hemoglobin difference in patients with and without delirium (11.6 versus 12.2 respectively). Our longitudinal analysis validated our comorbidity analysis and provided an understanding of the time between diagnosis of a risk factor to the development of delirium. In the UCSF data, for instance, roughly 3% of patients with anemia went on to develop delirium within 10 years of the diagnosis while less than 0.5% of patients without anemia developed delirium in this timeframe (hazard ratio of 9.4). One study in 700 patients specifically undergoing lumbar spinal fusion found that perioperative anemia is a risk factor for developing delirium post-operation \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Our study is the first to show that a diagnosis of anemia may be a more generalizable risk factor for developing inpatient delirium. Further studies need to be done to understand whether treating anemia more aggressively will reduce development of delirium.\u003c/p\u003e \u003cp\u003eThe large number of patients also enabled us to dissect the diagnostic associations with delirium in a sex-stratified manner. Our findings point to potentially new biological insights into the pathophysiology of delirium that are sex dependent. For instance, male patients had a significant association with \u003cem\u003eClostridioides difficile\u003c/em\u003e infections, an infection that often develops after antibiotic use. This could point to either the \u003cem\u003eClostridioides difficile\u003c/em\u003e toxin and the downstream consequences of the infection being associated with delirium or the antibiotic that was used that led to the infection being associated with delirium. Meanwhile, \u003cem\u003eKlebsiella\u003c/em\u003e and \u003cem\u003eE. coli\u003c/em\u003e infection were enriched in female patients with delirium, organisms that most commonly cause urinary tract infections, a source of infection more common in females compared to males. Dementia, one of the known risk factors of delirium, is also known to have sex-biased features, including in prevalence, clinical progression, and neuropathological findings \u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Interestingly, our association studies found that subtypes of dementia were associated with delirium in a sex-dependent manner. Further studies will need to be done to understand whether this is due to the underlying sex differences in the prevalence of these dementia subtypes or sex differences in the pathophysiology of these dementias contributing to delirium.\u003c/p\u003e \u003cp\u003eThere are several limitations to our study that need further investigation. Given that illness-severity is a well-studied risk factor for delirium, we used several indirect metrics of health status, including number of inpatient admissions and number of comorbidities, to find an appropriately matched control cohort to our delirium cohort. In this way, we were able to compare groups with similar health statuses to extract more subtle risk factors that may be associated with delirium. Though we matched on the number of comorbidities to identify a control cohort with a similar health status, some comorbidities are more functionally debilitating and severe than others. Not all factors relating to delirium, or any condition of interest are captured or measured\u003c/p\u003e \u003cp\u003ewithin the EHR as the data reflects only what is assessed or entered by clinicians. Given our control group had more pregnancy-related diagnoses, this suggests that though the control group may have a similar frequency of healthcare utilization, they are overall healthier than the matched delirium cohort. Future studies could use other health status metrics, such as the Charlson Comorbidity Index, which was not available in the datasets we used, or the American Society of Anesthesiologists (ASA) physical status classification score \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. ASA was available in our UC-wide dataset and was significantly higher in patients with delirium compared to control patients (\u003cb\u003eSupplementary Table\u0026nbsp;4\u003c/b\u003e). The mean values, however, were 2.95 in control patients versus 3.05 in delirium patients, suggesting this difference is likely not clinically meaningful, given the scores are integer numbers.\u003c/p\u003e \u003cp\u003eDefining an accurate delirium diagnosis in the EHR is challenging. Detection and clinical diagnosis of delirium itself is often challenging and likely underdiagnosed \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. As a result, the published prevalence of delirium in older patients varies widely, with some estimates around 23% \u003csup\u003e30\u003c/sup\u003e while others as high as 88% in palliative care patients \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Prevalence of delirium in our study was lower than these estimates, likely due to our restricted definition of delirium only including one OMOP concept ID and including younger patients. This definition, however, likely missed patients that did not fit this exact delirium definition. Documentation of delirium is unfortunately inconsistent in the EHR even with a confirmed diagnosis, with one study finding that only 9 out of 25 cases coded by ICD-9 code \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. RASS scores were available for some of our patients in the UC-wide dataset and showed scores were different between patients with delirium and controls, suggesting our inclusion criteria is likely consistent with clinical metrics of delirium. We, however, did not use RASS as a way to include or exclude patient as it is not a tool specific to delirium but rather an overall metric of sedation status. Future studies could include more patients by identifying delirium symptoms in clinical notes or including patients who have been screened for delirium using clinical tools such as the 4 A\u0026rsquo;s Test (4AT) or the confusion assessment method (CAM) \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, which were not available in the datasets we used.\u003c/p\u003e \u003cp\u003eIn this study, the definition of sex in the EHR is likely a combination of sex assigned at birth, legal sex, and sex determined by the clinician. Documentation of gender identity remains a challenge in the EHR and only recently has garnered attention to provide more inclusive and affirmative health care for all patients \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Further studies will need to be done to understand whether the associations found in this study are different when taking into context biological sex versus patient-reported gender identity. Similarly, other social determinants of health were not taken into account in this study, such as the patient\u0026rsquo;s level of formal education, which can influence cognitive reserve, as well as family / caregiver dynamics.\u003c/p\u003e \u003cp\u003eOverall, our study demonstrates the powerful application of the EHR to study heterogeneous disease processes and to better understand the risk factors and outcomes of disease at both a large patient population scale and a longitudinal time course. These results not only confirmed some of the known risk factors of delirium but also generated several new clinical hypotheses that will need to be further investigated. Applications of these analyses on other data types such as medication history may also expand the potential risk factors for delirium and increase the predictive power for the condition. These findings could also help develop future modeling studies for predicting which patients will develop delirium to focus prevention efforts towards these patients. Understanding the impact of sex differences in delirium remains understudied, and this study points to the importance of doing sex-stratified analyses and the potentially interesting pathophysiology of delirium that interacts with sex.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eResource Availability:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eLead Contact:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eFurther information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Marina Sirota ([email protected]).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMaterials availability:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThis study did not generate new unique reagents.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eData and code availability:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThe UCSF EHR database is available to individuals affiliated with UCSF who can contact the UCSF\u0026rsquo;s Clinical and Translational Science Institute (CTSI) ([email protected]) or the UCSF\u0026rsquo;s Information Commons team for more information ([email protected]). The UC-wide EHR database is only available to UC researchers who have completed analyses in their respective UC first and have provided justification for scaling their analyses across UC health centers (more details at https://www.ucop.edu/uc-health/departments/center-for-data-driven-insights-and-innovations-cdi2.html or by contacting [email protected]. The code used for analysis will be made available on Github. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExperimental Model and Study Participation Details:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll analysis of UCSF and UC-wide EHR data was performed under the approval of the Institutional Review Boards. All clinical data were de-identified and written informed consent was waived by the institutions. Patient cohorts were identified using the UCSF de-identified and UC-wide HIPAA-compliant limited data set OMOP EHR databases. The UCSF dataset included over five million patients from January 1, 1982 to February 20, 2023 while the UC-wide dataset included over seven million patients from January 1, 2012 to April 19, 2023 from 4 sites (UC San Diego, UC Los Angeles, UC Irvine, and UC Davis). Patients with inpatient delirium were identified using the OMOP concept ID 373995, SNOMED code 2776000 (corresponding to \u0026ldquo;Delirium\u0026rdquo;), filtered for first-time diagnosis of delirium during an inpatient stay (i.e., \u0026lsquo;visit of interest\u0026rsquo;). The inpatient control cohort with no delirium diagnosis was identified through propensity score (PS) matching (matchit R\u003csup\u003e37\u003c/sup\u003e) by a generalized linear model at a 1:1 ratio using a nearest neighbor method and the following matching criteria: assigned sex, patient-reported race, estimated age at admission, years in EHR prior to visit, total number of comorbidities and inpatient visits prior to the visit of interest, stay length, stay type (ICU vs non-ICU), death during admission, and UC location (for UC-wide dataset). Propensity score matching was used similar to other previously published works using clinical datasets \u003csup\u003e17, 23, 38, 39\u003c/sup\u003e. \u0026nbsp;For sex stratification, we utilized reported sex and excluded nonbinary and other categories due to low sample size, similar to a prior study \u003csup\u003e17\u003c/sup\u003e. See Table 1 for cohort demographic details.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethod Details:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDifferential comorbidity analysis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eComorbidities were identified prior to the visit of interest with the earliest entry of every diagnosis. All patients and their comorbidities were visualized using Uniform Manifold Approximation and Projection (UMAP) using the R-implementation of the umap-learn package from Python\u0026nbsp;\u003csup\u003e40\u003c/sup\u003e. UMAP is a form of dimensionality reduction to plot high-dimensional data into a lower-dimension. In this case, a table was created with all distinct comorbidities as columns and all patients as rows, with each table entry corresponding to 0 as the patient not having that diagnosis and 1 as the patient having that diagnosis. This table was then visualized as a UMAP with each point corresponding to one patient and encapsulating the similarity of that patient\u0026rsquo;s comorbidities to other patients based on distance. Correlations between variables (delirium status or sex) and UMAP coordinates were analyzed using Mann-Whitney \u003cem\u003eU\u003c/em\u003e-tests. Differential comorbidity analysis between patients with delirium and controls was done using Fisher\u0026rsquo;s exact test and significance determined by Bonferroni-corrected threshold of p-value \u0026lt; 0.05. ICD10-CM blocks were also used to visualize differential comorbidity results. The OMOP concept table was used to map the SNOMED codes to ICD10 codes then mapped to categorical modules using the ICD10 codes. Overlaps between results of analyses using UCSF versus UC-wide datasets were done using hypergeometric test or Spearman correlation. A similar analysis was done for the sex-stratified analysis.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDifferential laboratory test results analysis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLaboratory test results were collected for tests done prior to the visit of interest, and the median values for all numerical lab tests was calculated. Lab tests that had more than 95% of patients missing results were excluded from the analysis, adapted from a prior study\u0026nbsp;\u003csup\u003e17\u003c/sup\u003e. Lab value distributions were compared using Mann-Whitney \u003cem\u003eU\u003c/em\u003e-tests across delirium status or sex.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLongitudinal analyses:\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTime-to-event analysis was done by first identifying patients with a diagnosis of bipolar disorder (BD) and no prior delirium diagnosis. PS-matched control cohorts with no diagnosis of BD were identified using similar matching-criteria as described above. Time-to-event was calculated as time from first-time diagnosis of BD (or another non-BD diagnosis for the control patients) to either first inpatient delirium diagnosis, death, or loss to follow-up. Cox regression model was used to determine the hazard ratio, confidence intervals and significance (survival R\u003csup\u003e41\u003c/sup\u003e). Time-to-event analysis for mortality after delirium was done using a similar strategy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuantification and Statistical Analysis:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analyses were performed using R Statistical Software version 4.3.2 (R Core Team 2023) and the specific statistical tests used for each experiment are outlined in detail above in each respective section as well as in the figure legends respectively.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank the UCSF Information Commons for their help navigating the UCSF EHR database, especially Evan Phelps, and the UC Health Data Warehouse team for their help with the UC-wide database, especially Robert Follett. This work was supported by the following grants: National Institutes of Health grant T32GM007618 (L.K., A.T.), National Institutes of Health grant R01AG060393 (M.S.), National Institutes of Health grant R01AG057683 (M.S.)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: L.K., M.S.\u003c/p\u003e\n\u003cp\u003eMethodology: L.K., S.W., A.T., Y.L., J.K., I.E.A., T.O., M.S.\u003c/p\u003e\n\u003cp\u003eInvestigation: L.K., M.S.\u003c/p\u003e\n\u003cp\u003eVisualization: L.K.\u003c/p\u003e\n\u003cp\u003eFunding acquisition: L.K., M.S.\u003c/p\u003e\n\u003cp\u003eProject administration: L.K., M.S.\u003c/p\u003e\n\u003cp\u003eSupervision: M.S.\u003c/p\u003e\n\u003cp\u003eWriting \u0026ndash; original draft: L.K., M.S.\u003c/p\u003e\n\u003cp\u003eWriting \u0026ndash; review \u0026amp; editing: L.K., S.W., A.T., Y.L., J.K., I.E.A., T.O., E.R., M.S.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAssociation AP (2022) \u003cem\u003eDiagnostic and statistical manual of mental disorders (5th ed., text rev.)\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilson JE et al (2020) Delirium. Nat Rev Dis Primers 6:90\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWitlox J et al (2010) Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. JAMA 304:443\u0026ndash;451\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFiest KM et al (2021) Long-Term Outcomes in ICU Patients with Delirium: A Population-based Cohort Study. Am J Respir Crit Care Med 204:412\u0026ndash;420\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoldberg TE et al (2020) Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurol 77:1373\u0026ndash;1381\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurry L et al (2018) Antipsychotics for treatment of delirium in hospitalised non-ICU patients. Cochrane Database Syst Rev 6:Cd005594\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurton JK et al (2021) Non-pharmacological interventions for preventing delirium in hospitalised non-ICU patients. Cochrane Database Syst Rev 7:Cd013307\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLindroth H et al (2018) Systematic review of prediction models for delirium in the older adult inpatient. BMJ Open 8:e019223\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaker M et al (2006) Mutations in progranulin cause tau-negative frontotemporal dementia linked to chromosome 17. Nature 442:916\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePendlebury ST et al (2016) Delirium risk stratification in consecutive unselected admissions to acute medicine: validation of externally derived risk scores. Age Ageing 45:60\u0026ndash;65\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWong A et al (2018) Development and Validation of an Electronic Health Record-Based Machine Learning Model to Estimate Delirium Risk in Newly Hospitalized Patients Without Known Cognitive Impairment. JAMA Netw Open 1:e181018\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDouglas VC et al (2013) The AWOL tool: derivation and validation of a delirium prediction rule. J Hosp Med 8:493\u0026ndash;499\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSol\u0026agrave;-Miravete E et al (2018) Nursing assessment as an effective tool for the identification of delirium risk in older in-patients: A case-control study. J Clin Nurs 27:345\u0026ndash;354\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLopez-Lee C, Kodama L, Gan L (2022) Sex Differences in Neurodegeneration: The Role of the Immune System in Humans. Biol Psychiatry 91:72\u0026ndash;80\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrzepacz PT et al (2018) Delirium Phenotype by Age and Sex in a Pooled Data Set of Adult Patients. J Neuropsychiatry Clin Neurosci 30:294\u0026ndash;301\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWiredu K et al (2023) Sex Differences in the Incidence of Postoperative Delirium After Cardiac Surgery: A Pooled Analyses of Clinical Trials. Anesthesiology\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang AS et al (2022) Deep phenotyping of Alzheimer's disease leveraging electronic medical records identifies sex-specific clinical associations. Nat Commun 13:675\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi L et al (2015) Identification of type 2 diabetes subgroups through topological analysis of patient similarity. Sci Transl Med 7:311ra174\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbraham A et al (2022) Dense phenotyping from electronic health records enables machine learning-based prediction of preterm birth. BMC Med 20:333\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eM\u0026aring;nsson R, Joffe MM, Sun W, Hennessy S (2007) On the estimation and use of propensity scores in case-control and case-cohort studies. Am J Epidemiol 166:332\u0026ndash;339\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRose S, Laan MJ (2009) Why match? Investigating matched case-control study designs with causal effect estimation. Int J Biostat 5:1\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeberpals J et al (2021) Deep Learning-based Propensity Scores for Confounding Control in Comparative Effectiveness Research: A Large-scale, Real-world Data Study. Epidemiology 32:378\u0026ndash;388\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoldemariam SR, Tang AS, Oskotsky TT, Yaffe K, Sirota M (2023) Similarities and differences in Alzheimer's dementia comorbidities in racialized populations identified from electronic medical records. Commun Med (Lond) 3:50\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuguet N et al (2020) Using Electronic Health Records in Longitudinal Studies: Estimating Patient Attrition. Med Care 58(Suppl 1):S46\u0026ndash;s52\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEly EW et al (2003) Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 289:2983\u0026ndash;2991\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan JH et al (2015) The Diagnostic Performance of the Richmond Agitation Sedation Scale for Detecting Delirium in Older Emergency Department Patients. Acad Emerg Med 22:878\u0026ndash;882\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGold C et al (2022) Risk factors for delirium in elderly patients after lumbar spinal fusion. Clin Neurol Neurosurg 219:107318\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHurwitz EE et al (2017) Adding Examples to the ASA-Physical Status Classification Improves Correct Assignment to Patients. Anesthesiology 126:614\u0026ndash;622\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelirium is prevalent in (2019) older hospital inpatients and associated with adverse outcomes: results of a prospective multi-centre study on World Delirium Awareness Day. BMC Med 17:229\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibb K et al (2020) The consistent burden in published estimates of delirium occurrence in medical inpatients over four decades: a systematic review and meta-analysis study. Age Ageing 49:352\u0026ndash;360\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHosie A, Davidson PM, Agar M, Sanderson CR, Phillips J (2013) Delirium prevalence, incidence, and implications for screening in specialist palliative care inpatient settings: a systematic review. Palliat Med 27:486\u0026ndash;498\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHope C et al (2014) Documentation of delirium in the VA electronic health record. BMC Res Notes 7:208\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBellelli G et al (2014) Validation of the 4AT, a new instrument for rapid delirium screening: a study in 234 hospitalised older people. Age Ageing 43:496\u0026ndash;502\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInouye SK et al (1990) Clarifying confusion: the confusion assessment method. A new method for detection of delirium. Ann Intern Med 113:941\u0026ndash;948\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatel K, Lyon ME, Luu HS (2021) Providing Inclusive Care for Transgender Patients: Capturing Sex and Gender in the Electronic Medical Record. J Appl Lab Med 6:210\u0026ndash;218\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLau F, Antonio M, Davison K, Queen R, Devor A (2020) A rapid review of gender, sex, and sexual orientation documentation in electronic health records. J Am Med Inf Assoc 27:1774\u0026ndash;1783\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHo DI, King K, Stuart G (2011) MatchIt: Nonparametric Preprocessing for Parametric Causal Inference. J Stat Softw 42:1\u0026ndash;28\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarić I et al (2021) Decreased Mortality Rate Among COVID-19 Patients Prescribed Statins: Data From Electronic Health Records in the US. Front Med (Lausanne) 8:639804\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOskotsky T et al (2021) Mortality Risk Among Patients With COVID-19 Prescribed Selective Serotonin Reuptake Inhibitor Antidepressants. JAMA Netw Open 4:e2133090\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcInnes LH, Melville J (2020) J. UMAP: uniform manifold approximation and projection for dimension reduction. \u003cem\u003earXiv:1802.03426\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTherneau TMG (2000) P. M. \u003cem\u003eModeling Survival Data: Extending the Cox Model\u003c/em\u003eNew York\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4765249/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4765249/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDelirium is a detrimental mental condition often seen in older, hospitalized patients and is currently hard to predict. In this study, we leverage electronic health records (EHR) to identify 7,492 UCSF patients and 19,417 UC health system patients with an inpatient delirium diagnosis and the same number of control patients without delirium. We found significant associations between comorbidities or laboratory values and an inpatient delirium diagnosis, including metabolic abnormalities and psychiatric diagnoses. Some associations were sex-specific, including dementia subtypes and infections. We further explored the associations with anemia and bipolar disorder by conducting longitudinal analyses from the time of first diagnosis to development of delirium, demonstrating a significant relationship across time. Finally, we show that an inpatient delirium diagnosis leads to increased risk of mortality. These results demonstrate the powerful application of the EHR to shed insights into prior diagnoses and laboratory values that could help predict development of inpatient delirium and the importance of sex when making these assessments.\u003c/p\u003e","manuscriptTitle":"Sex-stratified analyses of comorbidities associated with an inpatient delirium diagnosis using real world data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-23 12:53:41","doi":"10.21203/rs.3.rs-4765249/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-medicine","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsmed","sideBox":"Learn more about [Communications Medicine](http://www.nature.com/commsmed)","snPcode":"43856","submissionUrl":"https://mts-commsmed.nature.com/cgi-bin/main.plex","title":"Communications Medicine","twitterHandle":"@commsmedicine","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a42f0bbf-f995-44bc-8ff9-6f3c02ba7376","owner":[],"postedDate":"July 23rd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":34940722,"name":"Health sciences/Diseases/Neurological disorders/Disorders of consciousness"},{"id":34940724,"name":"Health sciences/Diseases/Neurological disorders/Dementia"}],"tags":[],"updatedAt":"2025-07-23T07:12:02+00:00","versionOfRecord":{"articleIdentity":"rs-4765249","link":"https://doi.org/10.1038/s43856-025-00986-5","journal":{"identity":"communications-medicine","isVorOnly":false,"title":"Communications Medicine"},"publishedOn":"2025-07-22 04:00:00","publishedOnDateReadable":"July 22nd, 2025"},"versionCreatedAt":"2024-07-23 12:53:41","video":"","vorDoi":"10.1038/s43856-025-00986-5","vorDoiUrl":"https://doi.org/10.1038/s43856-025-00986-5","workflowStages":[]},"version":"v1","identity":"rs-4765249","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4765249","identity":"rs-4765249","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}