Your browser doesn't support javascript.
loading
Show: 20 | 50 | 100
Results 1 - 18 de 18
Filter
1.
Nat Genet ; 53(5): 630-637, 2021 05.
Article in English | MEDLINE | ID: mdl-33958779

ABSTRACT

The kidney is an organ of key relevance to blood pressure (BP) regulation, hypertension and antihypertensive treatment. However, genetically mediated renal mechanisms underlying susceptibility to hypertension remain poorly understood. We integrated genotype, gene expression, alternative splicing and DNA methylation profiles of up to 430 human kidneys to characterize the effects of BP index variants from genome-wide association studies (GWASs) on renal transcriptome and epigenome. We uncovered kidney targets for 479 (58.3%) BP-GWAS variants and paired 49 BP-GWAS kidney genes with 210 licensed drugs. Our colocalization and Mendelian randomization analyses identified 179 unique kidney genes with evidence of putatively causal effects on BP. Through Mendelian randomization, we also uncovered effects of BP on renal outcomes commonly affecting patients with hypertension. Collectively, our studies identified genetic variants, kidney genes, molecular mechanisms and biological pathways of key relevance to the genetic regulation of BP and inherited susceptibility to hypertension.


Subject(s)
Genetic Predisposition to Disease , Genomics , Hypertension/genetics , Kidney/pathology , Alternative Splicing/genetics , Blood Pressure/genetics , DNA Methylation/genetics , Genetic Variation , Genome-Wide Association Study , Humans , Mendelian Randomization Analysis , Polymorphism, Single Nucleotide/genetics , Quantitative Trait Loci/genetics
3.
Nat Genet ; 52(12): 1314-1332, 2020 12.
Article in English | MEDLINE | ID: mdl-33230300

ABSTRACT

Genetic studies of blood pressure (BP) to date have mainly analyzed common variants (minor allele frequency > 0.05). In a meta-analysis of up to ~1.3 million participants, we discovered 106 new BP-associated genomic regions and 87 rare (minor allele frequency ≤ 0.01) variant BP associations (P < 5 × 10-8), of which 32 were in new BP-associated loci and 55 were independent BP-associated single-nucleotide variants within known BP-associated regions. Average effects of rare variants (44% coding) were ~8 times larger than common variant effects and indicate potential candidate causal genes at new and known loci (for example, GATA5 and PLCB3). BP-associated variants (including rare and common) were enriched in regions of active chromatin in fetal tissues, potentially linking fetal development with BP regulation in later life. Multivariable Mendelian randomization suggested possible inverse effects of elevated systolic and diastolic BP on large artery stroke. Our study demonstrates the utility of rare-variant analyses for identifying candidate genes and the results highlight potential therapeutic targets.


Subject(s)
Blood Pressure/genetics , Gene Frequency/genetics , Genetic Predisposition to Disease/genetics , Hypertension/genetics , GATA5 Transcription Factor/genetics , Genome-Wide Association Study , Genotype , Humans , Mutation/genetics , Phospholipase C beta/genetics , Polymorphism, Single Nucleotide/genetics
4.
Kidney Int ; 95(3): 624-635, 2019 03.
Article in English | MEDLINE | ID: mdl-30784661

ABSTRACT

Nephrons scar and involute during aging, increasing the risk of chronic kidney disease. Little is known, however, about genetic mechanisms of kidney aging. We sought to define the signatures of age on the renal transcriptome using 563 human kidneys. The initial discovery analysis of 260 kidney transcriptomes from the TRANScriptome of renaL humAn TissuE Study (TRANSLATE) and the Cancer Genome Atlas identified 37 age-associated genes. For 19 of those genes, the association with age was replicated in 303 kidney transcriptomes from the Nephroseq resource. Surveying 42 nonrenal tissues from the Genotype-Tissue Expression project revealed that, for approximately a fifth of the replicated genes, the association with age was kidney-specific. Seventy-three percent of the replicated genes were associated with functional or histological parameters of age-related decline in kidney health, including glomerular filtration rate, glomerulosclerosis, interstitial fibrosis, tubular atrophy, and arterial narrowing. Common genetic variants in four of the age-related genes, namely LYG1, PPP1R3C, LTF and TSPYL5, correlated with the trajectory of age-related changes in their renal expression. Integrative analysis of genomic, epigenomic, and transcriptomic information revealed that the observed age-related decline in renal TSPYL5 expression was determined both genetically and epigenetically. Thus, this study revealed robust molecular signatures of the aging kidney and new regulatory mechanisms of age-related change in the kidney transcriptome.


Subject(s)
Aging/genetics , Nephrons/pathology , Renal Insufficiency, Chronic/genetics , Transcriptome/genetics , Adult , Aged , Aged, 80 and over , Aging/pathology , Computational Biology , DNA Methylation/genetics , Epigenomics , Female , Gene Expression Profiling , Genetic Variation , Glomerular Filtration Rate/physiology , Humans , Intracellular Signaling Peptides and Proteins/genetics , Lactoferrin/genetics , Male , Middle Aged , Muramidase/genetics , Nephrons/physiopathology , Nuclear Proteins/genetics , RNA-Seq , Renal Insufficiency, Chronic/pathology , Renal Insufficiency, Chronic/physiopathology
5.
Nat Commun ; 10(1): 29, 2019 01 03.
Article in English | MEDLINE | ID: mdl-30604766

ABSTRACT

Chronic kidney disease (CKD) affects ~10% of the global population, with considerable ethnic differences in prevalence and aetiology. We assemble genome-wide association studies of estimated glomerular filtration rate (eGFR), a measure of kidney function that defines CKD, in 312,468 individuals of diverse ancestry. We identify 127 distinct association signals with homogeneous effects on eGFR across ancestries and enrichment in genomic annotations including kidney-specific histone modifications. Fine-mapping reveals 40 high-confidence variants driving eGFR associations and highlights putative causal genes with cell-type specific expression in glomerulus, and in proximal and distal nephron. Mendelian randomisation supports causal effects of eGFR on overall and cause-specific CKD, kidney stone formation, diastolic blood pressure and hypertension. These results define novel molecular mechanisms and putative causal genes for eGFR, offering insight into clinical outcomes and routes to CKD treatment development.


Subject(s)
Glomerular Filtration Rate/genetics , Hypertension/genetics , Kidney Calculi/genetics , Kidney/physiopathology , Renal Insufficiency, Chronic/genetics , Adult , Aged , Blood Pressure/genetics , Ethnicity/genetics , Female , Genetic Loci/genetics , Genome-Wide Association Study , Histone Code/genetics , Histones/metabolism , Humans , Hypertension/ethnology , Hypertension/physiopathology , Kidney Calculi/ethnology , Kidney Calculi/physiopathology , Male , Middle Aged , Polymorphism, Single Nucleotide , Renal Insufficiency, Chronic/ethnology , Renal Insufficiency, Chronic/physiopathology
6.
Nat Commun ; 9(1): 4800, 2018 11 22.
Article in English | MEDLINE | ID: mdl-30467309

ABSTRACT

Genome-wide association studies (GWAS) have identified >100 loci of chronic kidney disease-defining traits (CKD-dt). Molecular mechanisms underlying these associations remain elusive. Using 280 kidney transcriptomes and 9958 gene expression profiles from 44 non-renal tissues we uncover gene expression partners (eGenes) for 88.9% of CKD-dt GWAS loci. Through epigenomic chromatin segmentation analysis and variant effect prediction we annotate functional consequences to 74% of these loci. Our colocalisation analysis and Mendelian randomisation in >130,000 subjects demonstrate causal effects of three eGenes (NAT8B, CASP9 and MUC1) on estimated glomerular filtration rate. We identify a common alternative splice variant in MUC1 (a gene responsible for rare Mendelian form of kidney disease) and observe increased renal expression of a specific MUC1 mRNA isoform as a plausible molecular mechanism of the GWAS association signal. These data highlight the variants and genes underpinning the associations uncovered in GWAS of CKD-dt.


Subject(s)
Genetic Predisposition to Disease/genetics , Genome-Wide Association Study , Polymorphism, Single Nucleotide , Renal Insufficiency, Chronic/genetics , Gene Expression Profiling , Genotype , Humans , Kidney/metabolism , Kidney/pathology , Phenotype , Renal Insufficiency, Chronic/pathology
7.
Nat Commun ; 9(1): 4792, 2018 11 12.
Article in English | MEDLINE | ID: mdl-30420621

ABSTRACT

In the original version of the Article, the gene symbol for tissue factor pathway inhibitor was inadvertently given as 'TFP1' instead of 'TFPI'. This has now been corrected in both the PDF and HTML versions of the Article.

8.
Nat Commun ; 9(1): 2526, 2018 06 28.
Article in English | MEDLINE | ID: mdl-29955040

ABSTRACT

Long-range chromosomal interactions bring distal regulatory elements and promoters together to regulate gene expression in biological processes. By performing promoter capture Hi-C (PCHi-C) on human embryonic stem cell-derived cardiomyocytes (hESC-CMs), we show that such promoter interactions are a key mechanism by which enhancers contact their target genes after hESC-CM differentiation from hESCs. We also show that the promoter interactome of hESC-CMs is associated with expression quantitative trait loci (eQTLs) in cardiac left ventricular tissue; captures the dynamic process of genome reorganisation after hESC-CM differentiation; overlaps genome-wide association study (GWAS) regions associated with heart rate; and identifies new candidate genes in such regions. These findings indicate that regulatory elements in hESC-CMs identified by our approach control gene expression involved in ventricular conduction and rhythm of the heart. The study of promoter interactions in other hESC-derived cell types may be of utility in functional investigation of GWAS-associated regions.


Subject(s)
Actinin/genetics , Calpain/genetics , Gene Regulatory Networks , Human Embryonic Stem Cells/metabolism , Myocytes, Cardiac/metabolism , Promoter Regions, Genetic , Actinin/metabolism , Calpain/metabolism , Cell Differentiation , Cell Line , Enhancer Elements, Genetic , Genome, Human , Genome-Wide Association Study , Heart Conduction System/cytology , Heart Conduction System/metabolism , Heart Rate/physiology , Heart Ventricles/cytology , Heart Ventricles/metabolism , Histones/genetics , Histones/metabolism , Human Embryonic Stem Cells/cytology , Humans , Myocytes, Cardiac/cytology , Protein Interaction Mapping , Protein Isoforms/genetics , Protein Isoforms/metabolism , Quantitative Trait Loci
9.
Kidney Blood Press Res ; 43(1): 55-67, 2018.
Article in English | MEDLINE | ID: mdl-29402805

ABSTRACT

Ageing of the kidney is a multi-dimensional process that occurs simultaneously at the molecular, cellular, histological, anatomical and physiological level. Nephron number and renal cortical volume decline, renal tubules become atrophic and glomeruli become sclerotic with age. These structural changes are accompanied by a decline in glomerular filtration rate, decreased sodium reabsorption and potassium excretion, reduced urinary concentrating capacity and alterations in the endocrine activity of the kidney. However, the pace of progression of these changes is not identical in everyone - individuals of the same age and seemingly similar clinical profile often exhibit stark differences in the age-related decline in renal health. Thus, chronological age poorly reflects the time-dependent changes that occur in the kidney. An ideal measure of renal vitality is biological kidney age - a measure of the age-related changes in physiological function. Replacing chronological age with biological age could provide numerous clinical benefits including improved prognostic accuracy in renal transplantation, better stratification of risk and identification of those who are on a fast trajectory to an age-related drop in kidney health.


Subject(s)
Aging/physiology , Kidney/physiology , Humans , Kidney Glomerulus/pathology , Kidney Tubules/physiopathology , Nephrons/pathology
10.
Hypertension ; 70(5): 1042-1048, 2017 11.
Article in English | MEDLINE | ID: mdl-28847892

ABSTRACT

We hypothesized that screening for nonadherence to antihypertensive treatment using liquid chromatography-tandem mass spectrometry-based biochemical analysis of urine/serum has therapeutic applications in nonadherent hypertensive patients. A retrospective analysis of hypertensive patients attending specialist tertiary care centers was conducted in 2 European countries (United Kingdom and Czech Republic). Nonadherence to antihypertensive treatment was diagnosed using biochemical analysis of urine (United Kingdom) or serum (Czech Republic). These results were subsequently discussed with each patient, and data on follow-up clinic blood pressure (BP) measurements were collected from clinical files. Of 238 UK patients who underwent biochemical urine analysis, 73 were nonadherent to antihypertensive treatment. Their initial urinary adherence ratio (the ratio of detected to prescribed antihypertensive medications) increased from 0.33 (0-0.67) to 1 (0.67-1) between the first and the last clinic appointments. The observed increase in the urinary adherence ratio in initially nonadherent UK patients was associated with the improved BP control; by the last clinic appointment, systolic and diastolic BPs were ≈19.5 and 7.5 mm Hg lower than at baseline (P=0.001 and 0.009, respectively). These findings were further corroborated in 93 nonadherent hypertensive patients from Czech Republic-their average systolic and diastolic BPs dropped by ≈32.6 and 17.4 mm Hg, respectively (P<0.001), on appointments after the biochemical analysis. Our data show that nonadherent hypertensive patients respond to liquid chromatography-tandem mass spectrometry-based biochemical analysis with improved adherence and significant BP drop. Such repeated biochemical analyses should be considered as a therapeutic approach in nonadherent hypertensive patients.


Subject(s)
Antihypertensive Agents , Biomarkers , Blood Pressure/drug effects , Hypertension , Medication Adherence/psychology , Medication Therapy Management/standards , Adult , Aged , Antihypertensive Agents/analysis , Antihypertensive Agents/therapeutic use , Biomarkers/blood , Biomarkers/urine , Blood Pressure Determination/methods , Blood Pressure Determination/psychology , Chromatography, Liquid/methods , Czech Republic/epidemiology , Female , Humans , Hypertension/diagnosis , Hypertension/drug therapy , Hypertension/epidemiology , Hypertension/psychology , Male , Middle Aged , Outcome and Process Assessment, Health Care , Quality Improvement , United Kingdom/epidemiology
11.
Eur J Hum Genet ; 25(11): 1181-1188, 2017 11.
Article in English | MEDLINE | ID: mdl-28853720

ABSTRACT

The Y chromosome has long been considered a 'genetic wasteland' on a trajectory to completely disappear from the human genome. The perception of its physiological function was restricted to sex determination and spermatogenesis. These views have been challenged in recent times with the identification of multiple ubiquitously expressed Y-chromosome genes and the discovery of several unexpected associations between the Y chromosome, immune system and complex polygenic traits. The collected evidence suggests that the Y chromosome influences immune and inflammatory responses in men, translating into genetically programmed susceptibility to diseases with a strong immune component. Phylogenetic studies reveal that carriers of a common European lineage of the Y chromosome (haplogroup I) possess increased risk of coronary artery disease. This occurs amidst upregulation of inflammation and suppression of adaptive immunity in this Y lineage, as well as inferior outcomes in human immunodeficiency virus infection. From structural analysis and experimental data, the UTY (Ubiquitously Transcribed Tetratricopeptide Repeat Containing, Y-Linked) gene is emerging as a promising candidate underlying the associations between Y-chromosome variants and the immunity-driven susceptibility to complex disease. This review synthesises the recent structural, experimental and clinical insights into the human Y chromosome in the context of men's susceptibility to disease (with a particular emphasis on cardiovascular disease) and provides an overview of the paradigm shift in the perception of the Y chromosome.


Subject(s)
Chromosomes, Human, Y/genetics , Genetic Predisposition to Disease , Cardiovascular Diseases/genetics , Humans , Immunity/genetics , Male
12.
Hypertension ; 69(6): 1113-1120, 2017 06.
Article in English | MEDLINE | ID: mdl-28461599

ABSTRACT

Nonadherence to antihypertensive treatment is a critical contributor to suboptimal blood pressure control. There are limited and heterogeneous data on the risk factors for nonadherence because few studies used objective-direct diagnostic methods. We used high-performance liquid chromatography-tandem mass spectrometry of urine and serum to detect nonadherence and explored its association with the main demographic- and therapy-related factors in 1348 patients with hypertension from 2 European countries. The rates of nonadherence to antihypertensive treatment were 41.6% and 31.5% in the UK and Czech populations, respectively. Nonadherence was inversely related to age and male sex. Each increase in the number of antihypertensive medications led to 85% and 77% increase in nonadherence (P<0.001) in the UK and Czech populations, respectively. The odds of nonadherence to diuretics were the highest among 5 classes of antihypertensive medications (P≤0.005 in both populations). The predictive model for nonadherence, including age, sex, diuretics, and the number of prescribed antihypertensives, showed area under the curves of 0.758 and 0.710 in the UK and Czech populations, respectively. The area under the curves for the UK model tested on the Czech data and for the Czech model tested on UK data were calculated at 0.708 and 0.756, respectively. We demonstrate that the number and class of prescribed antihypertensives are modifiable risk factors for biochemically confirmed nonadherence to blood pressure-lowering therapy. Further development of discriminatory models incorporating these parameters might prove clinically useful in assessment of nonadherence in countries where biochemical analysis is unavailable.


Subject(s)
Antihypertensive Agents/therapeutic use , Diuretics/therapeutic use , Hypertension/diagnosis , Hypertension/drug therapy , Medication Adherence/statistics & numerical data , Age Factors , Aged , Blood Pressure Determination/methods , Cohort Studies , Czech Republic , Female , Humans , Male , Middle Aged , Patient Compliance/statistics & numerical data , Predictive Value of Tests , Retrospective Studies , Risk Assessment , Severity of Illness Index , Sex Factors , United Kingdom
14.
BMJ ; 351: h5501, 2015 Nov 03.
Article in English | MEDLINE | ID: mdl-26537416

ABSTRACT

STUDY QUESTION: What is the prevalence of different types of potentially hazardous prescribing in general practice in the United Kingdom, and what is the variation between practices? METHODS: A cross sectional study included all adult patients potentially at risk of a prescribing or monitoring error defined by a combination of diagnoses and prescriptions in 526 general practices contributing to the Clinical Practice Research Datalink (CPRD) up to 1 April 2013. Primary outcomes were the prevalence of potentially hazardous prescriptions of anticoagulants, anti-platelets, NSAIDs, ß blockers, glitazones, metformin, digoxin, antipsychotics, combined hormonal contraceptives, and oestrogens and monitoring by blood test less frequently than recommended for patients with repeated prescriptions of angiotensin converting enzyme inhibitors and loop diuretics, amiodarone, methotrexate, lithium, or warfarin. STUDY ANSWER AND LIMITATIONS: 49 927 of 949 552 patients at risk triggered at least one prescribing indicator (5.26%, 95% confidence interval 5.21% to 5.30%) and 21 501 of 182 721 (11.8%, 11.6% to 11.9%) triggered at least one monitoring indicator. The prevalence of different types of potentially hazardous prescribing ranged from almost zero to 10.2%, and for inadequate monitoring ranged from 10.4% to 41.9%. Older patients and those prescribed multiple repeat medications had significantly higher risks of triggering a prescribing indicator whereas younger patients with fewer repeat prescriptions had significantly higher risk of triggering a monitoring indicator. There was high variation between practices for some indicators. Though prescribing safety indicators describe prescribing patterns that can increase the risk of harm to the patient and should generally be avoided, there will always be exceptions where the indicator is clinically justified. Furthermore there is the possibility that some information is not captured by CPRD for some practices-for example, INR results in patients receiving warfarin. WHAT THIS STUDY ADDS: The high prevalence for certain indicators emphasises existing prescribing risks and the need for their appropriate consideration within primary care, particularly for older patients and those taking multiple medications. The high variation between practices indicates potential for improvement through targeted practice level intervention. FUNDING, COMPETING INTERESTS, DATA SHARING: National Institute for Health Research through the Greater Manchester Primary Care Patient Safety Translational Research Centre (grant No GMPSTRC-2012-1). Data from CPRD cannot be shared because of licensing restrictions.


Subject(s)
Family Practice/standards , Medical Errors/statistics & numerical data , Patient Safety/standards , Practice Patterns, Physicians'/standards , Prescription Drugs/therapeutic use , Adult , Aged , Confidence Intervals , Cross-Sectional Studies , Family Practice/statistics & numerical data , Female , Humans , Male , Middle Aged , Practice Patterns, Physicians'/statistics & numerical data , United Kingdom
15.
Stud Health Technol Inform ; 216: 696-700, 2015.
Article in English | MEDLINE | ID: mdl-26262141

ABSTRACT

We demonstrate the use of electronic records and repeated measures of risk factors therein, to enable deeper understanding of the relationship between the full longitudinal trajectory of risk factors and outcomes. To illustrate, dynamic mixed effect modelling is used to summarise the level, trend and monitoring intensity of kidney function. The output from this model then forms covariates for a recurrent event Cox proportional hazards model for predicting adverse events (AE). Using data from Salford, UK, our multivariate model finds that steeper declines in kidney function raise the hazard of AE (HR: 1.13, 95% CI (1.05, 1.22)). There is a non-proportional relationship between the hazard of AE and the monitoring intensity of kidney function. Neither of these variables would be present in a classical risk prediction model.. This work illustrates the potential of using the full longitudinal profile of risk factors, rather than just their level. There is an opportunity for deep statistical learning leading to rich clinical insight using longitudinal signals in electronic data.


Subject(s)
Electronic Health Records/statistics & numerical data , Kidney Diseases/diagnosis , Kidney Diseases/mortality , Kidney Function Tests/statistics & numerical data , Models, Statistical , Proportional Hazards Models , Computer Simulation , Data Mining/methods , Data Mining/statistics & numerical data , Humans , Incidence , Longitudinal Studies , Prognosis , Reproducibility of Results , Risk Assessment/methods , Sensitivity and Specificity , Survival Analysis , United Kingdom/epidemiology
16.
Drug Saf ; 38(7): 671-82, 2015 Jul.
Article in English | MEDLINE | ID: mdl-26100143

ABSTRACT

INTRODUCTION: The extent of preventable medication-related hospital admissions and medication-related issues in primary care is significant enough to justify developing decision support systems for medication safety surveillance. The prerequisite for such systems is defining a relevant set of medication safety-related indicators and understanding the influence of both patient and general practice characteristics on medication prescribing and monitoring. OBJECTIVE: The aim of the study was to investigate the feasibility of linked primary and secondary care electronic health record data for surveillance of medication safety, examining not only prescribing but also monitoring, and associations with patient- and general practice-level characteristics. METHODS: A cross-sectional study was conducted using linked records of patients served by one hospital and over 50 general practices in Salford, UK. Statistical analysis consisted of mixed-effects logistic models, relating prescribing safety indicators to potential determinants. RESULTS: The overall prevalence (proportion of patients with at least one medication safety hazard) was 5.45 % for prescribing indicators and 7.65 % for monitoring indicators. Older patients and those on multiple medications were at higher risk of prescribing hazards, but at lower risk of missed monitoring. The odds of missed monitoring among all patients were 25 % less for males, 50 % less for patients in practices that provide general practitioner training, and threefold higher in practices serving the most deprived compared with the least deprived areas. Practices with more prescribing hazards did not tend to show more monitoring issues. CONCLUSIONS: Systematic collection, collation, and analysis of linked primary and secondary care records produce plausible and useful information about medication safety for a health system. Medication safety surveillance systems should pay close attention to patient age and polypharmacy with respect to both prescribing and monitoring failures; treat prescribing and monitoring as different statistical processes, rather than a combined measure of prescribing safety; and audit the socio-economic equity of missed monitoring.


Subject(s)
Adverse Drug Reaction Reporting Systems , Drug-Related Side Effects and Adverse Reactions/epidemiology , Electronic Health Records , Medical Record Linkage , Medication Errors/prevention & control , Prescription Drugs/adverse effects , Primary Health Care , Systems Integration , Adolescent , Adult , Age Factors , Aged , Aged, 80 and over , Cross-Sectional Studies , Drug-Related Side Effects and Adverse Reactions/diagnosis , Drug-Related Side Effects and Adverse Reactions/prevention & control , England/epidemiology , Feasibility Studies , Female , Humans , Logistic Models , Male , Middle Aged , Odds Ratio , Pharmacovigilance , Polypharmacy , Population Surveillance , Prescription Drugs/classification , Prevalence , Risk Assessment , Risk Factors , Secondary Care , Young Adult
17.
Burns ; 41(5): 925-34, 2015 Aug.
Article in English | MEDLINE | ID: mdl-25931158

ABSTRACT

INTRODUCTION: Predicting mortality from burn injury has traditionally employed logistic regression models. Alternative machine learning methods have been introduced in some areas of clinical prediction as the necessary software and computational facilities have become accessible. Here we compare logistic regression and machine learning predictions of mortality from burn. METHODS: An established logistic mortality model was compared to machine learning methods (artificial neural network, support vector machine, random forests and naïve Bayes) using a population-based (England & Wales) case-cohort registry. Predictive evaluation used: area under the receiver operating characteristic curve; sensitivity; specificity; positive predictive value and Youden's index. RESULTS: All methods had comparable discriminatory abilities, similar sensitivities, specificities and positive predictive values. Although some machine learning methods performed marginally better than logistic regression the differences were seldom statistically significant and clinically insubstantial. Random forests were marginally better for high positive predictive value and reasonable sensitivity. Neural networks yielded slightly better prediction overall. Logistic regression gives an optimal mix of performance and interpretability. DISCUSSION: The established logistic regression model of burn mortality performs well against more complex alternatives. Clinical prediction with a small set of strong, stable, independent predictors is unlikely to gain much from machine learning outside specialist research contexts.


Subject(s)
Burns/mortality , Decision Support Techniques , Machine Learning , Models, Statistical , Registries , Smoke Inhalation Injury/mortality , Adolescent , Adult , Age Factors , Bayes Theorem , Body Surface Area , Burns/pathology , Child , Child, Preschool , England , Female , Humans , Infant , Logistic Models , Male , Neural Networks, Computer , ROC Curve , Risk Assessment/methods , Software , Support Vector Machine , Wales , Young Adult
18.
Stud Health Technol Inform ; 192: 283-7, 2013.
Article in English | MEDLINE | ID: mdl-23920561

ABSTRACT

We compare the effectiveness of two types of verbal protocol, concurrent think aloud vs. retrospective sense making, for evaluating the usability of a clinical decision support tool. Thirty-five medical and nursing students participated in a usability experiment. Participants were asked to complete seven tasks using the system under evaluation. Eighteen students were allocated to the concurrent think aloud group and the remainder followed the retrospective protocol. The retrospective protocol was significantly more sensitive than the concurrent protocol in recording unique usability problems related to users' cognitive behaviour. These problems concerned the interpretation and comprehension of statistical output, search results and system messages. These findings can be explained by the retrospective protocol's greater ability to detect compound usability problems, capturing the cognitive dimensions of users' interactions with the interface in greater depth. Evaluations of clinical decision support systems should take an evidence-based approach to selecting protocols.


Subject(s)
Consumer Behavior/statistics & numerical data , Decision Support Systems, Clinical/statistics & numerical data , Reinforcement, Verbal , Software Validation , Software , User-Computer Interface , Algorithms , Comprehension , Female , Humans , Male , Retrospective Studies , United Kingdom
SELECTION OF CITATIONS
SEARCH DETAIL
...