Long-term cancer outcomes after bariatric surgery.

OBJECTIVE: Obesity is associated with increased cancer risk. Because of the substantial and sustained weight loss following bariatric surgery, postsurgical patients are ideal to study the association of weight loss and cancer. METHODS: Retrospectively (1982-2019), 21,837 bariatric surgery patients (surgery, 1982-2018) were matched 1:1 by age, sex, and BMI with a nonsurgical comparison group. Procedures included gastric bypass, gastric banding, sleeve gastrectomy, and duodenal switch. Primary outcomes included cancer incidence and mortality, stratified by obesity- and non-obesity-related cancers, sex, cancer stage, and procedure. RESULTS: Bariatric surgery patients had a 25% lower risk of developing any cancers compared with a nonsurgical comparison group (hazard ratio [HR] 0.75; 95% CI 0.69-0.81; p < 0.001). Cancer incidence was lower among female (HR 0.67; 95% CI 0.62-0.74; p < 0.001) but not male surgery patients, with the HR lower for females than for males (p < 0.001). Female surgery patients had a 41% lower risk for obesity-related cancers (i.e., breast, ovarian, uterine, and colon) compared with nonsurgical females (HR 0.59; 95% CI 0.52-0.66; p < 0.001). Cancer mortality was significantly lower after surgery in females (HR 0.53; 95% CI 0.44-0.64; p < 0.001). CONCLUSIONS: Bariatric surgery was associated with lower all-cancer and obesity-related cancer incidence among female patients. Cancer mortality was significantly lower among females in the surgical group versus the nonsurgical group.

Long-term all-cause and cause-specific mortality for four bariatric surgery procedures.

OBJECTIVE: This retrospective study incorporated long-term mortality results after different bariatric surgery procedures and for multiple age at surgery groups. METHODS: Participants with bariatric surgery (surgery) and without (non-surgery) were matched (1:1) for age, sex, BMI, and surgery date with a driver license application/renewal date. Mortality rates were compared by Cox regression, stratified by sex, surgery type, and age at surgery. RESULTS: Participants included 21,837 matched surgery and non-surgery pairs. Follow-up was up to 40 years (mean [SD], 13.2 [9.5] years). All-cause mortality was 16% lower in surgery compared with non-surgery groups (hazard ratio, 0.84; 95% CI: 0.79-0.90; p < 0.001). Significantly lower mortality after bariatric surgery was observed for both females and males. Mortality after surgery versus non-surgery decreased significantly by 29%, 43%, and 72% for cardiovascular disease, cancer, and diabetes, respectively. The hazard ratio for suicide was 2.4 times higher in surgery compared with non-surgery participants (95% CI: 1.57-3.68; p < 0.001), primarily in participants with ages at surgery between 18 and 34 years. CONCLUSIONS: Reduced all-cause mortality was durable for multiple decades, for multiple bariatric surgical procedures, for females and males, and for greater than age 34 years at surgery. Rate of death from suicide was significantly higher in surgery versus non-surgery participants only in the youngest age at surgery participants.

Creating a Real-World Data, United States Healthcare Claims-Based Adaptation of Kurtzke Functional Systems Scores for Assessing Multiple Sclerosis Severity and Progression.

INTRODUCTION: This article describes the development of a unique mapping of the Kurtzke Functional Systems Scores (KFSS) from International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM) codes among multiple sclerosis (MS) patients within a US Integrated Delivery Network (IDN). Valid identification of increasing disability may allow deeper insight into MS progression and possible treatments. METHODS: This cohort study identified MS patients in the IDN, Intermountain Healthcare. Experienced clinicians and informaticists mapped electronic health record ICD-9-CM codes to KFSS components generating a modified Kurtzke Expanded Disability Status Scale (EDSS). Modified EDSS scores were used to assess disability progression by calculating means, medians, ranges, and changes in KFSS and modified EDSS scores. RESULTS: Overall, 608/2960 (20.5%) patients were identified as having MS progression and presented a wide range of scores on the EDSS 10-point scale. The median (range) first and second EDSS scores were 0 (0-6) and 5 (1-8), respectively. The median (range) change from first to second score was 5 (1-7.5). The median first KFSS score for all systems was 0, and the mean differed among components. The highest mean first KFSS score (1.06) was measured for sensory function and lowest (0.12) for cerebellar functions. Of the 544 patients with their first EDSS scores in the ≤ 2.5 group, 75.2% and 15.1% had their second EDSS scores in group 3-5.5 and ≥ 6, respectively. Of the 62 patients with their first EDSS score in the 3-5.5 group, 58.1% had their second EDSS scores in group ≥ 6. CONCLUSION: This innovative mapping technique is a promising method for future comparative effectiveness and safety research of Disease-Modifying Therapy in Real-World Data repositories. Future research to validate and expand on this method in another healthcare database is encouraged.

Type 1 Diabetes incidence among youth in Utah: A geographical analysis.

Type 1 Diabetes (T1D) poses an increasing threat to public health, as incidence rates continue to rise globally. However, the etiology of T1D is still poorly understood, especially from the perspective of geography. The objective of this research is to examine the incidence of T1D among youth and to identify high-risk clusters and their association with socio-demographic and geographic variables. The study area was the entire state of Utah and included youth with T1D from birth to 19 years of age from 1998 to 2015 (n = 4161). Spatial clustering was measured both globally and locally using the Moran's I statistic and spatial scan statistic. Ordinary least squares (OLS) regression was used to measure the association of high-risk clusters with certain risk factors at the Census Block Group (CBG) level. The mean age at diagnosis was 9.3 years old. The mean incidence rate was 25.67 per 100,000 person-years (95% CI, 24.57-26.75). The incidence rate increased by 14%, from 23.94 per100,000 person-years in 1998 to 27.98 per 100,000 person-years in 2015, with an annual increase of 0.80%. The results of the spatial scan statistic found 42 high-risk clusters throughout the state. OLS regression analysis found a significant association with median household income, population density, and latitude. This study provides evidence that incidence rates of T1D are increasing annually in the state of Utah and that significant geographic high-risk clusters are associated with socio-demographic and geographic factors.

Nonalcoholic fatty liver disease: impact on healthcare resource utilization, liver transplantation and mortality in a large, integrated healthcare system.

BACKGROUND AND AIMS: NAFLD is the most prevalent liver disease globally, affecting 20% of the world population. Healthcare resource utilization (HRU) attributable to NAFLD has been difficult to define. METHODS: We performed a case control study on NAFLD patients from 2005 to 2015 in a large integrated healthcare system with an affiliated insurance company that prospectively captures HRU information. Outcomes encompassed costs, liver transplantation and mortality rates. RESULTS: There were 17,085 patients, of which 4512 were NAFLD cases and 12,573 were non-NAFLD controls. The cohorts were similar in age and gender distribution (p > 0.05). The NAFLD cohort had a younger mean age of death (60.9 vs. 63.3, p = 0.004) and had over twice the number of annual healthcare visits (14.6 vs. 7.1). The increased overall annual overall cost attributable to NAFLD (in 2015 $) was $449/year. Overall, NAFLD was independently associated with 17% higher annual attributable healthcare costs. More advanced NAFLD (FS 3-4) was associated with a 40% increase in median annual healthcare costs (vs. FS 0-2). The strongest predictors of HRU among patients with NAFLD were advanced fibrosis and medical co-morbidities. The rate of liver transplantation was 18 times greater (0.054%/year) in the NAFLD compared with the non-NAFLD cohort, while mortality rate was 1.7 times greater. CONCLUSIONS: Within a large, integrated healthcare system a diagnosis of NAFLD is independently associated with a 17% overall excess in HRU and a several-fold increase liver transplantation and mortality. Although the dollar amounts will change over time and between healthcare systems, the proportional need for HRU will have broad applicability and implications.

Incident infection following acute kidney injury with recovery to baseline creatinine: A propensity score matched analysis.

BACKGROUND: Severe acute kidney injury (AKI) is associated with subsequent infection. Whether AKI followed by a return to baseline creatinine is associated with incident infection is unknown. OBJECTIVE: We hypothesized that risk of both short and long term infection would be higher among patients with AKI and return to baseline creatinine than in propensity score matched peers without AKI in the year following a non-infectious hospital admission. DESIGN: Retrospective, propensity score matched cohort study. PARTICIPANTS: We identified 494 patients who were hospitalized between January 1, 1999 and December 31, 2009 and had AKI followed by return to baseline creatinine. These were propensity score matched to controls without AKI. MAIN MEASURES: The predictor variable was AKI defined by International Classification of Diseases, Ninth Revision (ICD-9) codes and by the Kidney Disease Improving Global Outcomes definition, with return to baseline creatinine defined as a decrease in serum creatinine level to within 10% of the baseline value within 7 days of hospital discharge. The outcome variable was incident infection defined by ICD-9 code within 1 year of hospital discharge. RESULTS: AKI followed by return to baseline creatinine was associated with a 4.5-fold increased odds ratio for infection (odds ratio 4.53 [95% CI, 2.43-8.45]; p<0.0001) within 30 days following discharge. The association between AKI and subsequent infection remained significant at 31-60 days and 91 to 365 days but not during 61-90 days following discharge. CONCLUSION: Among patients from an integrated health care delivery system, non-infectious AKI followed by return to baseline creatinine was associated with an increased odds ratio for infection in the year following discharge.

Percentage of Time in Range 70 to 139 mg/dL Is Associated With Reduced Mortality Among Critically Ill Patients Receiving IV Insulin Infusion.

BACKGROUND: In addition to hyperglycemia, hypoglycemia, and glycemic variability, reduced time in targeted blood glucose range (TIR) is associated with increased risk of death in critically ill patients. This relation between TIR and mortality may be confounded by diabetic status and antecedent glycemic control. METHODS: This study retrospectively analyzed critically ill patients managed with the same IV insulin protocol at multiple centers. The percentage of TIR between 70 and 139 mg/dL was calculated. Patients with diabetic ketoacidosis, patients who had < 10 blood glucose readings, and patients with repeat admissions were excluded. The highest recorded glycosylated hemoglobin value in the preceding 3 months or up to 1 month following admission were used as a surrogate for the patient's preexisting glucose control. Stratified regression analyses were performed for 30-day mortality, with covariates of age, sex, TIR ≥ 80%, Acute Physiology Score, and Charlson Comorbidity Index. RESULTS: A total of 9,028 patients, 53.2% of whom had diabetes, were studied. Median TIR was 84.1% for nondiabetic patients and 64.5% for patients with diabetes. Mortality was lower in those with TIR > 80% compared with those with TIR ≤ 80% (12.4% vs 19.2%; P < .001). TIR > 80% was independently associated with reduced mortality in nondiabetic patients (OR, 0.52; P < .001), patients with diabetes (OR, 0.69; P = .001), and patients with well-controlled disease (OR, 0.50; P < .001) but not in patients with poorly controlled disease (OR, 0.86; P = .40). CONCLUSIONS: TIR was independently associated with mortality in critically ill patients, particularly those with good antecedent glucose control.

CDS in a Learning Health Care System: Identifying Physicians' Reasons for Rejection of Best-Practice Recommendations in Pneumonia through Computerized Clinical Decision Support.

BACKGROUND: Local implementation of guidelines for pneumonia care is strongly recommended, but the context of care that affects implementation is poorly understood. In a learning health care system, computerized clinical decision support (CDS) provides an opportunity to both improve and track practice, providing insights into the implementation process. OBJECTIVES: This article examines physician interactions with a CDS to identify reasons for rejection of guideline recommendations. METHODS: We implemented a multicenter bedside CDS for the emergency department management of pneumonia that integrated patient data with guideline-based recommendations. We examined the frequency of adoption versus rejection of recommendations for site-of-care and antibiotic selection. We analyzed free-text responses provided by physicians explaining their clinical reasoning for rejection, using concept mapping and thematic analysis. RESULTS: Among 1,722 patient episodes, physicians rejected recommendations to send a patient home in 24%, leaving text in 53%; reasons for rejection of the recommendations included additional or alternative diagnoses beyond pneumonia, and comorbidities or signs of physiologic derangement contributing to risk of outpatient failure that were not processed by the CDS. Physicians rejected broad-spectrum antibiotic recommendations in 10%, leaving text in 76%; differences in pathogen risk assessment, additional patient information, concern about antibiotic properties, and admitting physician preferences were given as reasons for rejection. CONCLUSION: While adoption of CDS recommendations for pneumonia was high, physicians rejecting recommendations frequently provided feedback, reporting alternative diagnoses, additional individual patient characteristics, and provider preferences as major reasons for rejection. CDS that collects user feedback is feasible and can contribute to a learning health system.

Identifying Patients With Relapsing-Remitting Multiple Sclerosis Using Algorithms Applied to US Integrated Delivery Network Healthcare Data.

BACKGROUND: Relapsing-remitting multiple sclerosis (RRMS) has a major impact on affected patients; therefore, improved understanding of RRMS is important, particularly in the context of real-world evidence. OBJECTIVES: To develop and validate algorithms for identifying patients with RRMS in both unstructured clinical notes found in electronic health records (EHRs) and structured/coded health care claims data. METHODS: US Integrated Delivery Network data (2010-2014) were queried for study inclusion criteria (possible multiple sclerosis [MS] base cohort): one or more MS diagnosis code, patients aged 18 years or older, 1 year or more baseline history, and no other demyelinating diseases. Sets of algorithms were developed to search narrative text of unstructured clinical notes (EHR clinical notes-based algorithms) and structured/coded data (claims-based algorithms) to identify adult patients with RRMS, excluding patients with evidence of progressive MS. Medical records were reviewed manually for algorithm validation. Positive predictive value was calculated for both EHR clinical notes-based and claims-based algorithms. RESULTS: From a sample of 5308 patients with possible MS, 837 patients with RRMS were identified using only the EHR clinical notes-based algorithms and 2271 patients were identified using only the claims-based algorithms; 779 patients were identified using both algorithms. The positive predictive value was 99.1% (95% confidence interval [CI], 94.2%-100%) for the EHR clinical notes-based algorithms and 94.6% (95% CI, 89.1%-97.8%) to 94.9% (95% CI, 89.8%-97.9%) for the claims-based algorithms. CONCLUSIONS: The algorithms evaluated in this study identified a real-world cohort of patients with RRMS without evidence of progressive MS that can be studied in clinical research with confidence.

Risk of adverse maternal and fetal outcomes during pregnancy in living kidney donors: A matched cohort study.

BACKGROUND: We examined the risk of adverse pregnancy outcomes in primiparous kidney donors compared to matched controls. METHODS: Fifty-nine women with a history of kidney donation prior to their first pregnancy with normal renal function and no history of kidney disease, diabetes or chronic hypertension were matched 1:4 by age (within 2 years) and race to women with two kidneys using data from an integrated healthcare delivery system. Adverse pregnancy outcomes were defined as preterm delivery (delivery <37 weeks), delivery via cesarean section, gestational hypertension, preeclampsia/eclampsia, gestational diabetes, length of stay in the hospital >3 days, infant death/transfer to acute facility and low birthweight (<2500 g). RESULTS: Living kidney donors did not have a higher risk of adverse outcomes compared to matched controls. There was a trend toward an increased risk of preeclampsia/eclampsia in kidney donors but it did not reach statistical significance (Odds ratio [OR]: 2.96, 95% CI: 0.98-8.94, P = 0.06). However, in kidney donors ≤30 years of age, there was a fourfold increased risk of preeclampsia/eclampsia (OR: 4.09, 95% CI: 1.07-15.59, P = 0.04). CONCLUSION: Overall, the risk of pregnancy-associated complications following kidney donation is small but potential female kidney donors should be counseled on the possible increased risk of preeclampsia.

Association of Unilateral Renal Agenesis With Adverse Outcomes in Pregnancy: A Matched Cohort Study.

BACKGROUND: Data regarding the effect of a solitary kidney during pregnancy have come from studies of living kidney donors. We evaluated the risk for adverse pregnancy outcomes in women with a single kidney from renal agenesis. STUDY DESIGN: Matched cohort study. SETTING & PARTICIPANTS: Using data from 7,079 childbirths from an integrated health care delivery system from 1996 through 2015, we identified births from women with renal agenesis. Only first pregnancies and singleton births were included. After excluding those with diabetes and kidney disease, 200 women with renal agenesis were matched 1:4 by age (within 2 years), race, and history of hypertension to women with 2 kidneys. PREDICTOR: Renal agenesis defined by International Classification of Diseases, Ninth Revision (ICD-9) codes prior to pregnancy. OUTCOMES: The primary outcome was adverse maternal outcomes, including preterm delivery, delivery by cesarean section, preeclampsia/eclampsia, and hospital length of stay. Adverse neonatal end points were considered as a secondary outcome and included low birth weight (<2,500g) and infant death/transfer to acute inpatient facility. RESULTS: Mean gestational age at delivery was 37.9±2.1 weeks for women with renal agenesis compared to 38.6±1.8 weeks for women with 2 kidneys. Compared with women with 2 kidneys, those with renal agenesis had increased risk for preterm delivery (OR, 2.88; 95% CI, 1.86-4.45), delivery by cesarean section (OR, 2.11; 95% CI, 1.49-2.99), preeclampsia/eclampsia (OR, 2.41; 95% CI, 1.23-4.72), and length of stay longer than 3 days (OR, 1.81; 95% CI, 1.18-2.78). Renal agenesis was not significantly associated with increased risk for infant death/transfer to acute facility (OR, 2.60; 95% CI, 0.57-11.89) or low birth weight after accounting for preterm delivery (OR, 2.11; 95% CI, 0.76-5.88). LIMITATIONS: Renal agenesis was identified by ICD-9 code, not by imaging of the abdomen. CONCLUSION: Women with unilateral renal agenesis have a higher risk for adverse outcomes in pregnancy.

Clinical factors influencing the decision to order red blood cell transfusions for a sample of US dialysis patients
.

AIM: To characterize the clinical context for the decision to order red blood cell (RBC) transfusions in dialysis patients. MATERIALS AND METHODS: Retrospective review of medical records from three integrated health systems serving chronic dialysis patients. Subjects were randomly selected from all patients who received at least one transfusion between January 2009 and December 2013. Data abstracted included transfusion setting, prescribing clinician type, patient demographics and hemoglobin (Hb) concentration prior to transfusion, and cataloguing and prioritizing of clinical factors for their contribution to the decision to transfuse. Data from one system were stratified between transfusions before and after the 2011 dialysis payment reform and anemia drug label changes. RESULTS: Charts for 590 patients were reviewed. The primary reason for transfusion was low Hb (51%), medical conditions (22%), symptoms of anemia (18%), surgery-related (6%), and undetermined (3%). In 93% of cases, multiple factors were cited as contributors to the transfusion decision. Mean Hb prior to transfusion was 7.2 g/dL in patients where low Hb was the primary reason for transfusion (range: 4.0 - 9.9 g/dL). CONCLUSIONS: The decision to transfuse dialysis patients is influenced by multiple patient factors and medical conditions, of which low Hb is the main contributor to this decision about half of the time.
.

Inflammatory Bowel Disease Phenotype in Pediatric Primary Sclerosing Cholangitis.

BACKGROUND: Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease that most often occurs in association with inflammatory bowel disease (IBD). We examined whether the activity or colonic distribution of IBD differed in pediatric patients with and without PSC. METHODS: We compared colonic disease distribution, physician global assessment scores, Mayo endoscopic severity scores, IBD-related hospital admissions, and colonic resection surgery rate in a retrospective cohort of pediatric patients with IBD with and without PSC. RESULTS: We identified 37 patients with PSC-IBD, and 137 non-PSC matched IBD controls. Pancolitis was seen in 89.7 versus 72.4% (P = 0.051) of patients with PSC-IBD and rectal sparing in 24.3 versus 21.6% (P = 0.721) of patients with IBD. Physician global assessment and Mayo scores at presentation and in follow-up were similar in PSC-IBD and IBD. Patients with PSC-IBD had 0.19 admissions per person-year compared with 0.25 in patients with IBD. The incidence rate ratio for admission was 0.75 (95% confidence interval (CI), 0.51-1.08). The 5-year probability of colonic surgery was 16.4% (95% CI, 7.0-36.0) in patients with PSC-IBD and 24.7% (95% CI, 17.7-33.8) in patients with IBD (P = 0.271). In a multivariate model, male sex (hazard ratio [HR] = 2.2 [95% CI, 1.1-4.3]) and the presence of a non-PSC immune-mediated comorbidity {HR = 3.9 (95% CI, 1.5-10.4), but not PSC (HR = 0.5 [95% CI, 0.2-1.3])} or Crohn's disease (HR = 0.5 [95% CI, 0.1-1.5]), were risk factors for colonic surgery in pediatric IBD. CONCLUSIONS: Patients with IBD and PSC were more likely to present with pancolitis, but had similar rates of rectal sparing. Patients with IBD showed similar disease activity across a wide range of measures, at presentation and in follow-up, regardless of the presence of PSC.

Kidney disease and maternal and fetal outcomes in pregnancy.

BACKGROUND: Pregnancy in kidney disease is considered high risk, but the degree of this risk is unclear. We tested the hypothesis that kidney disease in pregnancy is associated with adverse maternal and fetal outcomes. STUDY DESIGN: Retrospective study comparing pregnant women with and without kidney disease. SETTING & PARTICIPANTS: Using data from an integrated health care delivery system from 2000 through 2013, a total of 778 women met the criteria for kidney disease. Using a pool of 74,105 women without kidney disease, we selected 778 women to use for matches for the women with kidney disease. These women were matched 1:1 by age, race, and history of diabetes, chronic hypertension, liver disease, and connective tissue disease. PREDICTOR: Kidney disease was defined using the NKF-KDOQI definition for chronic kidney disease or International Classification of Diseases, Ninth Revision codes prior to pregnancy or serum creatinine level > 1.2mg/dL and/or proteinuria in the first trimester. OUTCOMES & MEASUREMENTS: Maternal outcomes included preterm delivery, delivery by cesarean section, preeclampsia/eclampsia, length of stay at hospital (>3 days), and maternal death. Fetal outcomes included low birth weight (weight < 2,500g), small for gestational age, number of admissions to neonatal intensive care unit, and infant death. RESULTS: Compared with women without kidney disease, those with kidney disease had 52% increased odds of preterm delivery (OR, 1.52; 95% CI, 1.16-1.99) and 33% increased odds of delivery by cesarean section (OR, 1.33; 95% CI, 1.06-1.66). Infants born to women with kidney disease had 71% increased odds of admission to the neonatal intensive care unit or infant death compared with infants born to women without kidney disease (OR, 1.71; 95% CI, 1.17-2.51). Kidney disease also was associated with 2-fold increased odds of low birth weight (OR, 2.38; 95% CI, 1.64-3.44). Kidney disease was not associated with increased risk of maternal death. LIMITATIONS: Data for level of kidney function and cause of death not available. CONCLUSIONS: Kidney disease in pregnancy is associated independently with adverse maternal and fetal outcomes when other comorbid conditions are controlled by matching.

Coefficient of glucose variation is independently associated with mortality in critically ill patients receiving intravenous insulin.

INTRODUCTION: Both patient- and context-specific factors may explain the conflicting evidence regarding glucose control in critically ill patients. Blood glucose variability appears to correlate with mortality, but this variability may be an indicator of disease severity, rather than an independent predictor of mortality. We assessed blood glucose coefficient of variation as an independent predictor of mortality in the critically ill. METHODS: We used eProtocol-Insulin, an electronic protocol for managing intravenous insulin with explicit rules, high clinician compliance, and reproducibility. We studied critically ill patients from eight hospitals, excluding patients with diabetic ketoacidosis and patients supported with eProtocol-insulin for < 24 hours or with < 10 glucose measurements. Our primary clinical outcome was 30-day all-cause mortality. We performed multivariable logistic regression, with covariates of age, gender, glucose coefficient of variation (standard deviation/mean), Charlson comorbidity score, acute physiology score, presence of diabetes, and occurrence of hypoglycemia < 60 mg/dL. RESULTS: We studied 6101 critically ill adults. Coefficient of variation was independently associated with 30-day mortality (odds ratio 1.23 for every 10% increase, P < 0.001), even after adjustment for hypoglycemia, age, disease severity, and comorbidities. The association was higher in non-diabetics (OR = 1.37, P < 0.001) than in diabetics (OR 1.15, P = 0.001). CONCLUSIONS: Blood glucose variability is associated with mortality and is independent of hypoglycemia, disease severity, and comorbidities. Future studies should evaluate blood glucose variability.

Vitamin D level and risk of community-acquired pneumonia and sepsis.

Previous research has reported reduced serum 25-hydroxyvitamin D (25(OH)D) levels is associated with acute infectious illness. The relationship between vitamin D status, measured prior to acute infectious illness, with risk of community-acquired pneumonia (CAP) and sepsis has not been examined. Community-living individuals hospitalized with CAP or sepsis were age-, sex-, race-, and season-matched with controls. ICD-9 codes identified CAP and sepsis; chest radiograph confirmed CAP. Serum 25(OH)D levels were measured up to 15 months prior to hospitalization. Regression models adjusted for diabetes, renal disease, and peripheral vascular disease evaluated the association of 25(OH)D levels with CAP or sepsis risk. A total of 132 CAP patients and controls were 60 ± 17 years, 71% female, and 86% Caucasian. The 25(OH)D levels <37 nmol/L (adjusted odds ratio (OR) 2.57, 95% CI 1.08-6.08) were strongly associated with increased odds of CAP hospitalization. A total of 422 sepsis patients and controls were 65 ± 14 years, 59% female, and 91% Caucasian. The 25(OH)D levels <37 nmol/L (adjusted OR 1.75, 95% CI 1.11-2.77) were associated with increased odds of sepsis hospitalization. Vitamin D status was inversely associated with risk of CAP and sepsis hospitalization in a community-living adult population. Further clinical trials are needed to evaluate whether vitamin D supplementation can reduce risk of infections, including CAP and sepsis.

Ontology-based tools to expedite predictive model construction.

Large amounts of medical data are collected electronically during the course of caring for patients using modern medical information systems. This data presents an opportunity to develop clinically useful tools through data mining and observational research studies. However, the work necessary to make sense of this data and to integrate it into a research initiative can require substantial effort from medical experts as well as from experts in medical terminology, data extraction, and data analysis. This slows the process of medical research. To reduce the effort required for the construction of computable, diagnostic predictive models, we have developed a system that hybridizes a medical ontology with a large clinical data warehouse. Here we describe components of this system designed to automate the development of preliminary diagnostic models and to provide visual clues that can assist the researcher in planning for further analysis of the data behind these models.

Primary sclerosing cholangitis, autoimmune hepatitis, and overlap in Utah children: epidemiology and natural history.

UNLABELLED: The epidemiology and natural history of pediatric primary sclerosing cholangitis (PSC), autoimmune sclerosing cholangitis (ASC), and autoimmune hepatitis (AIH) are not well characterized. Using multiple, overlapping search strategies followed by a detailed records review, we identified all cases of pediatric PSC, ASC, AIH, and inflammatory bowel disease (IBD) in a geographically isolated region of the United States. We identified 607 cases of IBD, 29 cases of PSC, 12 cases of ASC, and 44 cases of AIH. The mean age at diagnosis was 13.0 years for PSC, 11.3 years for ASC, and 9.8 years for AIH. The incidence and prevalence of PSC, ASC, and AIH were 0.2 and 1.5 cases, 0.1 and 0.6 cases, and 0.4 and 3.0 cases per 100,000 children, respectively. The mean duration of follow-up was 5.9 years. The probability of developing complicated liver disease within 5 years of the diagnosis of liver disease was 37% [95% confidence interval (CI) = 21%-58%] for PSC, 25% (95% CI = 7%-70%) for ASC, and 15% (95% CI = 7%-33%) for AIH. The 5-year survival rates with the native liver were 78% (95% CI = 54%-91%) for PSC, 90% (95% CI = 47%-99%) for ASC, and 87% (95% CI = 71%-95%) for AIH. Cholangiocarcinoma developed in 2 of the 29 PSC patients (6.9%). PSC occurred in 9.9% of patients with ulcerative colitis (UC) and in 0.6% of patients with Crohn's disease (CD). ASC occurred in 2.3% of UC patients and 0.9% of CD patients. AIH occurred in 0.4% of UC patients and in 0.3% of CD patients. Liver disease occurred in 39 of 607 IBD patients (6.4%) overall. CONCLUSION: Immune-mediated liver diseases are important sources of morbidity in children. Using a population-based design, this study quantifies the burden and natural history of immune-mediated liver disease in children.

An ontology-driven, diagnostic modeling system.

OBJECTIVES: To present a system that uses knowledge stored in a medical ontology to automate the development of diagnostic decision support systems. To illustrate its function through an example focused on the development of a tool for diagnosing pneumonia. MATERIALS AND METHODS: We developed a system that automates the creation of diagnostic decision-support applications. It relies on a medical ontology to direct the acquisition of clinic data from a clinical data warehouse and uses an automated analytic system to apply a sequence of machine learning algorithms that create applications for diagnostic screening. We refer to this system as the ontology-driven diagnostic modeling system (ODMS). We tested this system using samples of patient data collected in Salt Lake City emergency rooms and stored in Intermountain Healthcare's enterprise data warehouse. RESULTS: The system was used in the preliminary development steps of a tool to identify patients with pneumonia in the emergency department. This tool was compared with a manually created diagnostic tool derived from a curated dataset. The manually created tool is currently in clinical use. The automatically created tool had an area under the receiver operating characteristic curve of 0.920 (95% CI 0.916 to 0.924), compared with 0.944 (95% CI 0.942 to 0.947) for the manually created tool. DISCUSSION: Initial testing of the ODMS demonstrates promising accuracy for the highly automated results and illustrates the route to model improvement. CONCLUSIONS: The use of medical knowledge, embedded in ontologies, to direct the initial development of diagnostic computing systems appears feasible.