A unidirectional mapping of ICD-8 to ICD-10 codes, for harmonized longitudinal analysis of diseases.

Periodic revisions of the international classification of diseases (ICD) ensure that the classification reflects new practices and knowledge; however, this complicates retrospective research as diagnoses are coded in different versions. For longitudinal disease trajectory studies, a crosswalk is an essential tool and a comprehensive mapping between ICD-8 and ICD-10 has until now been lacking. In this study, we map all ICD-8 morbidity codes to ICD-10 in the expanded Danish ICD version. We mapped ICD-8 codes to ICD-10, using a many-to-one system inspired by general equivalence mappings such that each ICD-8 code maps to a single ICD-10 code. Each ICD-8 code was manually and unidirectionally mapped to a single ICD-10 code based on medical setting and context. Each match was assigned a score (1 of 4 levels) reflecting the quality of the match and, if applicable, a "flag" signalling choices made in the mapping. We provide the first complete mapping of the 8596 ICD-8 morbidity codes to ICD-10 codes. All Danish ICD-8 codes representing diseases were mapped and 5106 (59.4%) achieved the highest consistency score. Only 334 (3.9%) of the ICD-8 codes received the lowest mapping consistency score. The mapping provides a scaffold for translation of ICD-8 to ICD-10, which enable longitudinal disease studies back to and 1969 in Denmark and to 1965 internationally with further adaption.

Language-agnostic pharmacovigilant text mining to elicit side effects from clinical notes and hospital medication records.

We sought to craft a drug safety signalling pipeline associating latent information in clinical free text with exposures to single drugs and drug pairs. Data arose from 12 secondary and tertiary public hospitals in two Danish regions, comprising approximately half the Danish population. Notes were operationalised with a fastText embedding, based on which we trained 10 270 neural-network models (one for each distinct single-drug/drug-pair exposure) predicting the risk of exposure given an embedding vector. We included 2 905 251 admissions between May 2008 and June 2016, with 13 740 564 distinct drug prescriptions; the median number of prescriptions was 5 (IQR: 3-9) and in 1 184 340 (41%) admissions patients used ≥5 drugs concomitantly. A total of 10 788 259 clinical notes were included, with 179 441 739 tokens retained after pruning. Of 345 single-drug signals reviewed, 28 (8.1%) represented possibly undescribed relationships; 186 (54%) signals were clinically meaningful. Sixteen (14%) of the 115 drug-pair signals were possible interactions, and two (1.7%) were known. In conclusion, we built a language-agnostic pipeline for mining associations between free-text information and medication exposure without manual curation, predicting not the likely outcome of a range of exposures but also the likely exposures for outcomes of interest. Our approach may help overcome limitations of text mining methods relying on curated data in English and can help leverage non-English free text for pharmacovigilance.

Using Machine Learning to Identify Patients at High Risk of Inappropriate Drug Dosing in Periods with Renal Dysfunction.

PURPOSE: Dosing of renally cleared drugs in patients with kidney failure often deviates from clinical guidelines, so we sought to elicit predictors of receiving inappropriate doses of renal risk drugs. PATIENTS AND METHODS: We combined data from the Danish National Patient Register and in-hospital data on drug administrations and estimated glomerular filtration rates for admissions between 1 October 2009 and 1 June 2016, from a pool of about 2.6 million persons. We trained artificial neural network and linear logistic ridge regression models to predict the risk of five outcomes (>0, ≥1, ≥2, ≥3 and ≥5 inappropriate doses daily) with index set 24 hours after admission. We used time-series validation for evaluating discrimination, calibration, clinical utility and explanations. RESULTS: Of 52,451 admissions included, 42,250 (81%) were used for model development. The median age was 77 years; 50% of admissions were of women. ≥5 drugs were used between admission start and index in 23,124 admissions (44%); the most common drug classes were analgesics, systemic antibacterials, diuretics, antithrombotics, and antacids. The neural network models had better discriminative power (all AUROCs between 0.77 and 0.81) and were better calibrated than their linear counterparts. The main prediction drivers were use of anti-inflammatory, antidiabetic and anti-Parkinson's drugs as well as having a diagnosis of chronic kidney failure. Sex and age affected predictions but slightly. CONCLUSION: Our models can flag patients at high risk of receiving at least one inappropriate dose daily in a controlled in-silico setting. A prospective clinical study may confirm that this holds in real-life settings and translates into benefits in hard endpoints.

Phenome-Wide Analysis of Short- and Long-Run Disease Incidence Following Recurrent Pregnancy Loss Using Data From a 39-Year Period.

Background It is unclear how recurrent pregnancy loss (RPL) impacts disease risk and whether there is a difference in risk between women with or without a live birth before RPL (primary versus secondary RPL). We investigated the disease risk following RPL, and whether there was a difference between primary and secondary RPL. Methods and Results Using population-wide healthcare registries from Denmark, we identified a cohort of 1 370 896 ever-pregnant women aged 12 to 40 years between 1977 and 2016. Of this cohort, 10 691 (0.77%) fulfilled the criteria for RPL (50.0% primary RPL). Average follow-up was 15.8 years. Incidence rate ratios were calculated in a phenome-wide manner. Diagnoses related to assessment and diagnosis of RPL and those appearing later in life were separated using a mixture model. Primary RPL increased the risk of subsequent cardiovascular disorders, including atherosclerosis, cerebral infarction, heart failure, and pulmonary embolism, as well as systemic lupus erythematosus, chronic obstructive pulmonary disease, anxiety, and obsessive-compulsive disorder. Women with secondary RPL had no increased risk of cardiovascular disorders. However, we observed an increased risk of gastrointestinal disorders such as irritable bowel syndrome and intestinal malabsorption, as well as mental disorders and obstetric complications. Conclusions RPL is a risk factor for a spectrum of disorders, which is different for primary and secondary RPL. Screening following RPL explains some associations, but the remaining findings suggest that RPL influences or shares cause with cardiovascular disorders, autoimmune disorders, and mental disorders. Research into the pathophysiology of RPL and later diseases merits further investigation.

Selective use of corifollitropin for controlled ovarian stimulation for IVF in patients with low anti-Müllerian hormone.

Corifollitropin, a long-acting follicle-stimulating hormone (FSH) analogue used for in vitro fertilization (IVF), does not allow individualization of dosage, and the ovarian response is similar to around 300 IU of daily recombinant FSH. This has raised concerns about the risk of ovarian hyperstimulation syndrome (OHSS) when used in standard patients. We administered corifollitropin selectively to patients with anticipated low to moderate ovarian response based on antimüllerian hormone levels in the lower quartile. The end points were oocyte distribution and occurrence of OHSS in women with AMH ≤15 pmol/L. The study included a cohort of 368 patients treated in 599 cycles. Post hoc the cohort was subdivided according to AMH. With increasing baseline AMH, the number of oocytes increased from a mean of 2.7 (range 0-8 with AMH <3 pmol/L) to 6.3 (range 0-15 with AMH 10-15 pmol/L) oocytes. Cancellations of retrievals and transfers decreased significantly with increasing AMH. Overall, the ongoing live pregnancy rate per started cycle was 15.2%. None developed OHSS. No cycles were cancelled or needed triggering of ovulation using a GnRH agonist due to risk of OHSS. Selective use of corifollitropin in patients with AMH in the lower quartile is a safe and appropriate way of optimising stimulation.

Role of Serotonin Transporter Changes in Depressive Responses to Sex-Steroid Hormone Manipulation: A Positron Emission Tomography Study.

BACKGROUND: An adverse response to acute and pronounced changes in sex-hormone levels during, for example, the perimenopausal or postpartum period appears to heighten risk for major depression in women. The underlying risk mechanisms remain elusive but may include transiently compromised serotonergic brain signaling. Here, we modeled a biphasic ovarian sex hormone fluctuation using a gonadotropin-releasing hormone agonist (GnRHa) and evaluated if emergence of depressive symptoms was associated with change in cerebral serotonin transporter (SERT) binding following intervention. METHODS: A double-blind, randomized, placebo-controlled study included 63 healthy female volunteers (mean age 24.3 ± 4.9 years) with regular menstrual cycles between 23 and 35 days. Participants were randomized to active (goserelin [GnRHa] 3.6 mg implant) or placebo intervention. Sixty women completed follow-up and entered the analyses. Primary outcome measures were changes from baseline in depressive symptoms assessed on the 17-item Hamilton Depression Rating Scale and SERT binding as imaged by [(11)C]DASB positron emission tomography. Outcome measures were acquired at baseline in the follicular phase (cycle day 6.6 ± 2.2) and at follow-up (16.2 ± 2.6 days after intervention start). RESULTS: Sex hormone manipulation with GnRHa significantly triggered subclinical depressive symptoms within-group (p = .003) and relative to placebo (p = .02), which were positively associated with net decreases in estradiol levels (p = .02) from baseline within the GnRHa group. Depressive symptoms were associated with increases in neocortical SERT binding in the GnRHa group relative to placebo (p = .003). CONCLUSIONS: Our data imply both serotonergic signaling and estradiol in the mechanisms by which sex-steroid hormone fluctuations provoke depressive symptoms and thus provide a rationale for future preventive strategies in high-risk groups.