In this study, we assessed to what extent data on the subject of TPPR (transfer, persistence, prevalence, recovery) that are obtained through an older STR typing kit can be used in an activity-level evaluation for a case profiled with a more modern STR kit. Newer kits generally hold more loci and may show higher sensitivity especially when reduced reaction volumes are used, and this could increase the evidential value at the source level. On the other hand, the increased genotyping information may invoke a higher number of contributors in the weight of evidence calculations, which could affect the evidential values as well. An activity scenario well explored in earlier studies [1,2] was redone using volunteers with known DNA profiles. DNA extracts were analyzed with three different approaches, namely using the optimal DNA input for (1) an older and (2) a newer STR typing system, and (3) using a standard, volume-based input combined with replicate PCR analysis with only the newer STR kit. The genotyping results were analyzed for various aspects such as percentage detected alleles and relative peak height contribution for background and the contributors known to be involved in the activity. Next, source-level LRs were calculated and the same trends were observed with regard to inclusionary and exclusionary LRs for persons who had or had not been in direct contact with the sampled areas. We subsequently assessed the impact on the outcome of the activity-level evaluation in an exemplary case by applying the assigned probabilities to a Bayesian network. We infer that data from different STR kits can be combined in the activity-level evaluations.
DOCUMENT
BACKGROUND: Conflicting evidence exists on the effectiveness of routinely measured vital signs on the early detection of increased probability of adverse events.PURPOSE: To assess the clinical relevance of routinely measured vital signs in medically and surgically hospitalized patients through a systematic review.DATA SOURCES: MEDLINE, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Cumulative Index to Nursing and Allied Health Literature, and Meta-analysen van diagnostisch onderzoek (in Dutch; MEDION) were searched to January 2013.STUDY SELECTION: Prospective studies evaluating routine vital sign measurements of hospitalized patients, in relation to mortality, septic or circulatory shock, intensive care unit admission, bleeding, reoperation, or infection.DATA EXTRACTION: Two reviewers independently assessed potential bias and extracted data to calculate likelihood ratios (LRs) and predictive values.DATA SYNTHESIS: Fifteen studies were performed in medical (n = 7), surgical (n = 4), or combined patient populations (n = 4; totaling 42,565 participants). Only three studies were relatively free from potential bias. For temperature, the positive LR (LR+) ranged from 0 to 9.88 (median 1.78; n = 9 studies); heart rate 0.82 to 6.79 (median 1.51; n = 5 studies); blood pressure 0.72 to 4.7 (median 2.97; n = 4 studies); oxygen saturation 0.65 to 6.35 (median 1.74; n = 2 studies); and respiratory rate 1.27 to 1.89 (n = 3 studies). Overall, three studies reported area under the Receiver Operator Characteristic (ROC) curve (AUC) data, ranging from 0.59 to 0.76. Two studies reported on combined vital signs, in which one study found an LR+ of 47.0, but in the other the AUC was not influenced.CONCLUSIONS: Some discriminative LR+ were found, suggesting the clinical relevance of routine vital sign measurements. However, the subject is poorly studied, and many studies have methodological flaws. Further rigorous research is needed specifically intended to investigate the clinical relevance of routinely measured vital signs.CLINICAL RELEVANCE: The results of this research are important for clinical nurses to underpin daily routine practices and clinical decision making.
DOCUMENT
Background: A pragmatic, stepped wedge trial design can be an appealing design to evaluate complex interventions in real-life settings. However, there are certain pitfalls that need to be considered. This paper reports on the experiences and lessons learned from the conduct of a cluster randomized, stepped wedge trial evaluating the effect of the Hospital Elder Life Program (HELP) in a Dutch hospital setting to prevent older patients from developing delirium. Methods: We evaluated our trial which was conducted in eight departments in two hospitals in hospitalized patients aged 70 years or older who were at risk for delirium by reflecting on the assumptions that we had and on what we intended to accomplish when we started, as compared to what we actually realized in the different phases of our study. Lessons learned on the design, the timeline, the enrollment of eligible patients and the use of routinely collected data are provided accompanied by recommendations to address challenges. Results: The start of the trial was delayed which caused subsequent time schedule problems. The requirement for individual informed consent for a quality improvement project made the inclusion more prone to selection bias. Most units experienced major difficulties in including patients, leading to excluding two of the eight units from participation. This resulted in failing to include a similar number of patients in the control condition versus the intervention condition. Data on outcomes routinely collected in the electronic patient records were not accessible during the study, and appeared to be often missing during analyses. Conclusions: The stepped wedge, cluster randomized trial poses specific risks in the design and execution of research in real-life settings of which researchers should be aware to prevent negative consequences impacting the validity of their results. Valid conclusions on the effectiveness of the HELP in the Dutch hospital setting are hampered by the limited quantity and quality of routine clinical data in our pragmatic trial. Executing a stepped wedge design in a daily practice setting using routinely collected data requires specific attention to ethical review, flexibility, a spacious time schedule, the availability of substantial capacity in the research team and early checks on the data availability and quality.
DOCUMENT
Objective: To automatically recognize self-acknowledged limitations in clinical research publications to support efforts in improving research transparency.Methods: To develop our recognition methods, we used a set of 8431 sentences from 1197 PubMed Central articles. A subset of these sentences was manually annotated for training/testing, and inter-annotator agreement was calculated. We cast the recognition problem as a binary classification task, in which we determine whether a given sentence from a publication discusses self-acknowledged limitations or not. We experimented with three methods: a rule-based approach based on document structure, supervised machine learning, and a semi-supervised method that uses self-training to expand the training set in order to improve classification performance. The machine learning algorithms used were logistic regression (LR) and support vector machines (SVM).Results: Annotators had good agreement in labeling limitation sentences (Krippendorff's α = 0.781). Of the three methods used, the rule-based method yielded the best performance with 91.5% accuracy (95% CI [90.1-92.9]), while self-training with SVM led to a small improvement over fully supervised learning (89.9%, 95% CI [88.4-91.4] vs 89.6%, 95% CI [88.1-91.1]).Conclusions: The approach presented can be incorporated into the workflows of stakeholders focusing on research transparency to improve reporting of limitations in clinical studies.
DOCUMENT
Het doel van dit onderzoek is te onderzoeken onder welke omstandigheden en onder welke condities relatief moderne modelleringstechnieken zoals support vector machines, neural networks en random forests voordelen zouden kunnen hebben in medisch-wetenschappelijk onderzoek en in de medische praktijk in vergelijking met meer traditionele modelleringstechnieken, zoals lineaire regressie, logistische regressie en Cox regressie.
MULTIFILE
Background: Modern modeling techniques may potentially provide more accurate predictions of dichotomous outcomes than classical techniques. Objective: In this study, we aimed to examine the predictive performance of eight modeling techniques to predict mortality by frailty. Methods: We performed a longitudinal study with a 7-year follow-up. The sample consisted of 479 Dutch community-dwelling people, aged 75 years and older. Frailty was assessed with the Tilburg Frailty Indicator (TFI), a self-report questionnaire. This questionnaire consists of eight physical, four psychological, and three social frailty components. The municipality of Roosendaal, a city in the Netherlands, provided the mortality dates. We compared modeling techniques, such as support vector machine (SVM), neural network (NN), random forest, and least absolute shrinkage and selection operator, as well as classical techniques, such as logistic regression, two Bayesian networks, and recursive partitioning (RP). The area under the receiver operating characteristic curve (AUROC) indicated the performance of the models. The models were validated using bootstrapping. Results: We found that the NN model had the best validated performance (AUROC=0.812), followed by the SVM model (AUROC=0.705). The other models had validated AUROC values below 0.700. The RP model had the lowest validated AUROC (0.605). The NN model had the highest optimism (0.156). The predictor variable “difficulty in walking” was important for all models. Conclusions: Because of the high optimism of the NN model, we prefer the SVM model for predicting mortality among community-dwelling older people using the TFI, with the addition of “gender” and “age” variables. External validation is a necessary step before applying the prediction models in a new setting.
DOCUMENT
Objective. There are no Dutch language disease-specific questionnaires for patients with patellofemoral pain syndrome available that could help Dutch physiotherapists to assess and monitor these symptoms and functional limitations. The aim of this study was to translate the original disease-specific Kujala Patellofemoral Score into Dutch and evaluate its reliability. Methods. The questionnaire was translated from English into Dutch in accordance with internationally recommended guidelines. Reliability was determined in 50 stable subjects with an interval of 1 week. The patient inclusion criteria were age between 14 and 60 years; knowledge of the Dutch language; and the presence of at least three of the following symptoms: pain while taking the stairs, pain when squatting, pain when running, pain when cycling, pain when sitting with knees flexed for a prolonged period, grinding of the patella and a positive clinical patella test. The internal consistency, test–retest reliability, measurement error and limits of agreement were calculated. Results. Internal consistency was 0.78 for the first assessment and 0.80 for the second assessment. The intraclass correlation coefficient (ICCagreement) between the first and second assessments was 0.98. The mean difference between the first and second measurements was 0.64, and standard deviation was 5.51. The standard error measurement was 3.9, and the smallest detectable change was 11. The Bland and Altman plot shows that the limits of agreement are 10.37 and 11.65. Conclusions. The results of the present study indicated that the test–retest reliability translated Dutch version of the Kujala Patellofemoral Score questionnaire is equivalent of the test– retest original English language version and has good internal consistency. Trial registration NTR (TC = 3258). Copyright © 2015 John Wiley & Sons, Ltd.
DOCUMENT
Machine learning models have proven to be reliable methods in classification tasks. However, little research has been done on classifying dwelling characteristics based on smart meter & weather data before. Gaining insights into dwelling characteristics can be helpful to create/improve the policies for creating new dwellings at NZEB standard. This paper compares the different machine learning algorithms and the methods used to correctly implement the models. These methods include the data pre-processing, model validation and evaluation. Smart meter data was provided by Groene Mient, which was used to train several machine learning algorithms. The models that were generated by the algorithms were compared on their performance. The results showed that Recurrent Neural Network (RNN) 2performed the best with 96% of accuracy. Cross Validation was used to validate the models, where 80% of the data was used for training purposes and 20% was used for testing purposes. Evaluation metrices were used to produce classification reports, which can indicate which of the models work the best for this specific problem. The models were programmed in Python.
DOCUMENT
Objective: Despite the increasing availability of eRehabilitation, its use remains limited. The aim of this study was to assess factors associated with willingness to use eRehabilitation. Design: Cross-sectional survey. Subjects: Stroke patients, informal caregivers, health-care professionals. Methods: The survey included personal characteristics, willingness to use eRehabilitation (yes/no) and barri-ers/facilitators influencing this willingness (4-point scale). Barriers/facilitators were merged into factors. The association between these factors and willingness to use eRehabilitation was assessed using logistic regression analyses. Results: Overall, 125 patients, 43 informal caregivers and 105 healthcare professionals participated in the study. Willingness to use eRehabilitation was positively influenced by perceived patient benefits (e.g. reduced travel time, increased motivation, better outcomes), among patients (odds ratio (OR) 2.68; 95% confidence interval (95% CI) 1.34–5.33), informal caregivers (OR 8.98; 95% CI 1.70–47.33) and healthcare professionals (OR 6.25; 95% CI 1.17–10.48). Insufficient knowledge decreased willingness to use eRehabilitation among pa-tients (OR 0.36, 95% CI 0.17–0.74). Limitations of the study include low response rates and possible response bias. Conclusion: Differences were found between patients/informal caregivers and healthcare professionals. Ho-wever, for both groups, perceived benefits of the use of eRehabilitation facilitated willingness to use eRehabili-tation. Further research is needed to determine the benefits of such programs, and inform all users about the potential benefits, and how to use eRehabilitation. Lay Abstract The use of digital eRehabilitation after stroke (e.g. in serious games, e-consultation and education) is increasing. However, the use of eRehabilitation in daily practice is limited. As a first step in increasing the use of eRehabilitation in stroke care, this study examined which factors influence the willingness of stroke patients, informal caregivers and healthcare professionals to use eRehabilitation. Beliefs about the benefits of eRehabilitation were found to have the largest positive impact on willingness to use eRehabilitation. These benefits included reduced travel time, increased adherence to therapy or motivation, and better health outcomes. The willingness to use eRehabilitation is limited by a lack of knowledge about how to use eRehabilitation.
MULTIFILE
Machine learning models have proven to be reliable methods in classification tasks. However, little research has been conducted on the classification of dwelling characteristics based on smart meter and weather data before. Gaining insights into dwelling characteristics, which comprise of the type of heating system used, the number of inhabitants, and the number of solar panels installed, can be helpful in creating or improving the policies to create new dwellings at nearly zero-energy standard. This paper compares different supervised machine learning algorithms, namely Logistic Regression, Support Vector Machine, K-Nearest Neighbor, and Long-short term memory, and methods used to correctly implement these algorithms. These methods include data pre-processing, model validation, and evaluation. Smart meter data, which was used to train several machine learning algorithms, was provided by Groene Mient. The models that were generated by the algorithms were compared on their performance. The results showed that the Long-short term memory performed the best with 96% accuracy. Cross Validation was used to validate the models, where 80% of the data was used for training purposes and 20% was used for testing purposes. Evaluation metrics were used to produce classification reports, which indicates that the Long-short term memory outperforms the compared models on the evaluation metrics for this specific problem.
DOCUMENT
