Prediction and risk stratification from hospital discharge records based on Hierarchical sLDA

Background The greatly accelerated development of information technology has conveniently provided adoption for risk stratification, which means more beneficial for both patients and clinicians. Risk stratification offers accurate individualized prevention and therapeutic decision making etc. Hospital discharge records (HDRs) routinely include accurate conclusions of diagnoses of the patients. For this reason, in this paper, we propose an improved model for risk stratification in a supervised fashion by exploring HDRs about coronary heart disease (CHD). Methods We introduced an improved four-layer supervised latent Dirichlet allocation (sLDA) approach called Hierarchical sLDA model, which categorized patient features in HDRs as patient feature-value pairs in one-hot way according to clinical guidelines for lab test of CHD. To address the data missing and imbalance problem, RFs and SMOTE methods are used respectively. After TF-IDF processing of datasets, variational Bayes expectation-maximization method and generalized linear model were used to recognize the latent clinical state of a patient, i.e., risk stratification, as well as to predict CHD. Accuracy, macro-F1, training and testing time performance were used to evaluate the performance of our model. Results According to the characteristics of our datasets, i.e., patient feature-value pairs, we construct a supervised topic model by adding one more Dirichlet distribution hyperparameter to sLDA. Compared with established supervised algorithm Multi-class sLDA model, we demonstrate that our proposed approach enhances training time by 59.74% and testing time by 25.58% but almost no loss of average prediction accuracy on our datasets. Conclusions A model for risk stratification and prediction of CHD based on sLDA model was proposed. Experimental results show that Hierarchical sLDA model we proposed is competitive in time performance and accuracy. Hierarchical processing of patient features can significantly improve the disadvantages of low efficiency and time-consuming Gibbs sampling of sLDA model.

Estimating human leishmaniasis burden in Spain using the capture-recapture method, 2016–2017

Leishmaniasis is endemic and a mandatory reporting disease in Spain since 1982. However, between 1996 and 2014, surveillance on public health was decentralized and only some autonomous regions monitored the disease. The aim of this study is to estimate the incidence of leishmaniasis and to evaluate the extent of underreporting in Spain. A capture-recapture (CRC) study was conducted to calculate the incidence of human leishmaniasis using reports from the National Surveillance Network (RENAVE) and the Hospital Discharge Records of the National Health System (CMBD) for 2016 and 2017. During the study period, 802 cases were reported to RENAVE and there were 1,149 incident hospitalizations related to leishmaniasis. The estimated incidence rates through the CRC study were 0.79 per 100,000 inhabitants for visceral leishmaniasis (VL), 0.88 (cutaneous leishmaniasis (CL)) and 0.12 (mucocutaneous leishmaniasis (MCL)) in 2016 and 0.86 (VL), 1.04 (CL) and 0.12 (MCL) in 2017. An underreporting of 14.7–20.2% for VL and 50.4–55.1% for CL was found. The CRC method has helped us to assess the sensitivity and representativeness of leishmaniasis surveillance in Spain, being a useful tool to assess whether the generalization of leishmaniasis surveillance throughout the Spanish territory achieves a reduction in underreporting.

Validation of algorithms for selecting rheumatoid arthritis patients in the Tuscan healthcare administrative databases

Validation of algorithms for selecting patients from healthcare administrative databases (HAD) is recommended. This PATHFINDER study section is aimed at testing algorithms to select rheumatoid arthritis (RA) patients from Tuscan HAD (THAD) and assessing RA diagnosis time interval between the medical chart date and that of THAD. A population was extracted from THAD. The information of the medical charts at the Rheumatology Unit of Pisa University Hospital represented the reference. We included first ever users of biologic disease modifying anti-rheumatic drugs (bDMARDs) between 2014 and 2016 (index date) with at least a specialist visit at the Rheumatology Unit of the Pisa University Hospital recorded from 2013 to the index date. Out of these, we tested four index tests (algorithms): (1) RA according to hospital discharge records or emergency department admissions (ICD-9 code, 714*); (2) RA according to exemption code from co-payment (006); (3) RA according to hospital discharge records or emergency department admissions AND RA according to exemption code from co-payment; (4) RA according to hospital discharge records or emergency department admissions OR RA according to exemption code from co-payment. We estimated sensitivity, specificity, positive and negative predicted values (PPV and NPV) with 95% confidence interval (95% CI) and the RA diagnosis median time interval (interquartile range, IQR). Two sensitivity analyses were performed. Among 277 reference patients, 103 had RA. The fourth algorithm identified 96 true RA patients, PPV 0.78 (95% CI 0.70–0.85), sensitivity 0.93 (95% CI 0.86–0.97), specificity 0.84 (95% CI 0.78–0.90), and NPV 0.95 (95% CI 0.91–0.98). The sensitivity analyses confirmed performance. The time measured between the actual RA diagnosis date recorded in medical charts and that assumed in THAD was 2.2 years (IQR 0.5–8.4). In conclusion, this validation showed the fourth algorithm as the best. The time interval elapsed between the actual RA diagnosis date in medical charts and that extrapolated from THAD has to be considered in the design of future studies.

The regional decline and rise of tick-borne encephalitis incidence do not correlate with Lyme borreliosis, Austria, 2005 to 2018

Background Tick-borne encephalitis (TBE) virus is a human pathogen that is expanding its endemic zones in Europe, emerging in previously unaffected regions. In Austria, increasing incidence in alpine regions in the west has been countered by a decline in traditional endemic areas to the east of the country. Aim To shed light on the cause of this disparity, we compared the temporal changes of human TBE incidences in all federal provinces of Austria with those of Lyme borreliosis (LB), which has the same tick vector and rodent reservoir. Methods This comparative analysis was based on the surveillance of hospitalised TBE cases by the National Reference Center for TBE and on the analysis of hospitalised LB cases from hospital discharge records across all of Austria from 2005 to 2018. Results The incidences of the two diseases and their annual fluctuations were not geographically concordant. Neither the decline in TBE in the eastern lowlands nor the increase in western alpine regions is paralleled by similar changes in the incidence of LB. Conclusion The discrepancy between changes in incidence of TBE and LB support the contributions of virus-specific factors beyond the mere availability of tick vectors and/or human outdoor activity, which are a prerequisite for the transmission of both diseases. A better understanding of parameters controlling human pathogenicity and the maintenance of TBE virus in its natural vector−host cycle will generate further insights into the focal nature of TBE and can potentially improve forecasts of TBE risk on smaller regional scales.

Spatial Patterns of Endometriosis Incidence. A Study in Friuli Venezia Giulia (Italy) in the Period 2004–2017

Background: Diagnosis of endometriosis and evaluation of incidence data are complex tasks because the disease is identified laparoscopically and confirmed histologically. Incidence estimates reported in literature are widely inconsistent, presumably reflecting geographical variability of risk and the difficulty of obtaining reliable data. Methods: We retrieved incident cases of endometriosis in women aged 15–50 years using hospital discharge records and pathology databases of the Friuli Venezia Giulia region in the calendar period 2004–2017. We studied the spatial pattern of endometriosis incidence applying Bayesian approaches to Disease Mapping, and profiled municipalities at higher risk controlling for multiple comparisons using both q-values and a fully Bayesian approach. Results: 4125 new cases of endometriosis were identified in the age range 15 to 50 years in the period 2004–2017. The incidence rate (x100 000) is 111 (95% CI 110–112), with a maximum of 160 in the age group 31–35 years. The geographical distribution of endometriosis incidence showed a very strong north-south spatial gradient. We consistently identified a group of five neighboring municipalities at higher risk (RR 1.31 95% CI 1.13; 1.52), even accounting for ascertainment bias. Conclusions: The cluster of 5 municipalities in the industrialized and polluted south-east part of the region is suggestive. However, due to the ecologic nature of the present study, information on the patients’ characteristics and exposure histories are limited. Individual studies, including biomonitoring, and life-course studies are necessary to better evaluate our findings.

Trends in Spontaneous and Medically Indicated Preterm Birth in Twins versus Singletons: A California Cohort 2007 to 2011

Objective The study aimed to describe preterm birth (PTB) rates, subtypes, and risk factors in twins compared with singletons to better understand reasons for the decline in PTB rate between 2007 and 2011. Study Design This was a retrospective population-based analysis using the California linked birth certificates and maternal-infant hospital discharge records from 2007 to 2011. The main outcomes were overall, spontaneous (following spontaneous labor or preterm premature rupture of membranes), and medically indicated PTB at various gestational age categories: <37, <32, and 34 to 36 weeks in twins and singletons. Results Among the 2,290,973 singletons and 28,937 twin live births pairs included, overall PTB <37 weeks decreased by 8.46% (6.77–6.20%) in singletons and 7.17% (55.31–51.35%) in twins during the study period. In singletons, this was primarily due to a 24.91% decrease in medically indicated PTB with almost no change in spontaneous PTB, whereas in twins indicated PTB declined 7.02% and spontaneous PTB by 7.39%. Conclusion Recent declines in PTB in singletons appear to be largely due to declines in indicated PTB, whereas both spontaneous and indicated PTB declined in twins. Key Points

Hospitalization for Short-Term Diabetes-Related Complications: Focus on Patients Aged over 85 Years

(1) Background: The prevalence of diabetes in elderly people is frequently high. When occurring in the elderly, diabetes is often accompanied by complications and comorbidities, at least one in 60% and four or more in 40% of older people with diabetes. As far as short-term complications among the elderly are concerned, hypoglycemia and hyperglycemic crises prove to be frequent. The aim of this study was to investigate the difference in hospitalization for short-term diabetes complications in patients below and over 85 years of age. (2) Methods: Data were collected from hospital discharge records (HDRs) of all hospital admissions that occurred in Abruzzo Region, Italy, from 2006 to 2015. Only diabetic patients aged over 65 years were included. Outcomes included were diabetic ketoacidosis, hyperosmolar coma, hypoglycemic shock, iatrogenic hypoglycemic coma, and other diabetic comas. (3) Results: During the study period, 144,376 admissions were collected, 116,305 (80.56%) of which referred to patients below 85 years. Those aged over 85 years were significantly associated to all short-term diabetes-related complications with the exception of ketoacidosis. (4) Conclusions: In older diabetic patients, the avoidance of short-term diabetes complications are a greater concern than in younger patients. Diabetes management among very elderly patients should be tailored accordingly to patient characteristics.