Search

Background: Better phenotyping of routinely collected coded data would be useful for research and health improvement. For example, the precision of coded data for hemorrhagic stroke (intracerebral hemorrhage [ICH] and subarachnoid hemorrhage [SAH]) may be as poor as < 50%. This work aimed to investigate the feasibility and added value of automated methods applied to clinical radiology reports to improve stroke subtyping.Methods: From a sub-population of 17,249 Scottish UK Biobank participants, we ascertained those with an incident stroke code in hospital, death record or primary care administrative data by September 2015, and ≥ 1 clinical brain scan report. We used a combination of natural language processing and clinical knowledge inference on brain scan reports to assign a stroke subtype (ischemic vs ICH vs SAH) for each participant and assessed performance by precision and recall at entity and patient levels.Results: Of 225 participants with an incident stroke code, 207 had a relevant brain scan report and were included in this study. Entity level precision and recall ranged from 78 to 100%. Automated methods showed precision and recall at patient level that were very good for ICH (both 89%), good for SAH (both 82%), but, as expected, lower for ischemic stroke (73%, and 64%, respectively), suggesting coded data remains the preferred method for identifying the latter stroke subtype.Conclusions: Our automated method applied to radiology reports provides a feasible, scalable and accurate solution to improve disease subtyping when used in conjunction with administrative coded health data. Future research should validate these findings in a different population setting. [ABSTRACT FROM AUTHOR]

Background: Ovarian cancer early detection markers CA125, CA15.3, HE4, and CA72.4 vary between healthy women, limiting their utility for screening. Methods: We evaluated cross-sectional relationships between lifestyle and reproductive factors and these markers among controls (n = 1910) from a nested case-control study in the European Prospective Investigation into Cancer and Nutrition (EPIC). Improvements in discrimination of prediction models adjusting for correlates of the markers were evaluated among postmenopausal women in the nested case-control study (n = 590 cases). Generalized linear models were used to calculate geometric means of CA125, CA15.3, and HE4. CA72.4 above vs. below limit of detection was evaluated using logistic regression. Early detection prediction was modeled using conditional logistic regression. Results: CA125 concentrations were lower, and CA15.3 higher, in post- vs. premenopausal women (p ≤ 0.02). Among postmenopausal women, CA125 was higher among women with higher parity and older age at menopause (ptrend ≤ 0.02), but lower among women reporting oophorectomy, hysterectomy, ever use of estrogen-only hormone therapy, or current smoking (p < 0.01). CA15.3 concentrations were higher among heavier women and in former smokers (p ≤ 0.03). HE4 was higher with older age at blood collection and in current smokers, and inversely associated with OC use duration, parity, and older age at menopause (≤ 0.02). No associations were observed with CA72.4. Adjusting for correlates of the markers in prediction models did not improve the discrimination. Conclusions: This study provides insights into sources of variation in ovarian cancer early detection markers in healthy women and informs about the utility of individualizing marker cutpoints based on epidemiologic factors. Electronic supplementary material The online version of this article (doi:10.1186/s13048-017-0315-6) contains supplementary material, which is available to authorized users.

Introduction: Associations of hormone-receptor positive breast cancer with excess adiposity are reasonably well characterized; however, uncertainty remains regarding the association of body mass index (BMI) with hormone-receptor negative malignancies, and possible interactions by hormone replacement therapy (HRT) use. Methods: Within the European EPIC cohort, Cox proportional hazards models were used to describe the relationship of BMI, waist and hip circumferences with risk of estrogen-receptor (ER) negative and progesterone-receptor (PR) negative (n = 1,021) and ER+PR+ (n = 3,586) breast tumors within five-year age bands. Among postmenopausal women, the joint effects of BMI and HRT use were analyzed. Results: For risk of ER-PR- tumors, there was no association of BMI across the age bands. However, when analyses were restricted to postmenopausal HRT never users, a positive risk association with BMI (third versus first tertile HR = 1.47 (1.01 to 2.15)) was observed. BMI was inversely associated with ER+PR+ tumors among women aged \(≤49 years (per 5 kg/m^2 increase, HR = 0.79 (95\%CI 0.68 to 0.91))\), and positively associated with risk among women ≥65 years (HR = 1.25 (1.16 to 1.34)). Adjusting for BMI, waist and hip circumferences showed no further associations with risks of breast cancer subtypes. Current use of HRT was significantly associated with an increased risk of receptor-negative (HRT current use compared to HRT never use HR: 1.30 (1.05 to 1.62)) and positive tumors (HR: 1.74 (1.56 to 1.95)), although this risk increase was weaker for ER-PR- disease \((P_{het} = 0.035)\). The association of HRT was significantly stronger in the leaner women \((BMI ≤22.5 kg/m^2)\) than for more overweight women \((BMI ≥25.9 kg/m^2)\) for, both, ER-PR- (HR: 1.74 (1.15 to 2.63)) and ER+PR+ (HR: 2.33 (1.84 to 2.92)) breast cancer and was not restricted to any particular HRT regime. Conclusions: An elevated BMI may be positively associated with risk of ER-PR- tumors among postmenopausal women who never used HRT. Furthermore, postmenopausal HRT users were at an increased risk of ER-PR- as well as ER+PR+ tumors, especially among leaner women. For hormone-receptor positive tumors, but not for hormone-receptor negative tumors, our study confirms an inverse association of risk with BMI among young women of premenopausal age. Our data provide evidence for a possible role of sex hormones in the etiology of hormone-receptor negative tumors.