Dataset Information

Repeatedly measured predictors: a comparison of methods for prediction modeling.

ABSTRACT:

Background

In literature, not much emphasis has been placed on methods for analyzing repeatedly measured independent variables, even less so for the use in prediction modeling specifically. However, repeated measurements could especially be interesting for the construction of prediction models. Therefore, our objective was to evaluate different methods to model a repeatedly measured independent variable and a long-term fixed outcome variable into a prediction model.

Methods

Six methods to handle a repeatedly measured predictor were applied to develop prediction models. Methods were evaluated with respect to the models' predictive quality (explained variance R ² and the area under the curve (AUC)) and their properties were discussed. The models included overweight and BMI-standard deviation score (BMI-SDS) at age 10 years as outcome and seven BMI-SDS measurements between 0 and 5.5 years as longitudinal predictor. Methods for comparison encompassed developing models including: all measurements; a single (here: the last) measurement; a mean or maximum value of all measurements; changes between subsequent measurements; conditional measurements; and growth curve parameters.

Results

All methods, except for using the maximum or mean, resulted in prediction models for overweight of similar predictive quality, with adjusted Nagelkerke R ² ranging between 0.230 and 0.244 and AUC ranging between 0.799 and 0.807. Continuous BMI-SDS prediction showed similar results.

Conclusions

The choice of method depends on hypothesized predictor-outcome associations, available data, and requirements of the prediction model. Overall, the growth curve method seems to be the most flexible method capable of incorporating longitudinal predictor information without loss in predictive quality.

SUBMITTER: Welten M

PROVIDER: S-EPMC6460730 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Repeatedly measured predictors: a comparison of methods for prediction modeling.

Welten Marieke M de Kroon Marlou L A MLA Renders Carry M CM Steyerberg Ewout W EW Raat Hein H Twisk Jos W R JWR Heymans Martijn W MW

Diagnostic and prognostic research 20180213

<h4>Background</h4>In literature, not much emphasis has been placed on methods for analyzing repeatedly measured independent variables, even less so for the use in prediction modeling specifically. However, repeated measurements could especially be interesting for the construction of prediction models. Therefore, our objective was to evaluate different methods to model a repeatedly measured independent variable and a long-term fixed outcome variable into a prediction model.<h4>Methods</h4>Six me ...[more]

PMID: 31093555

Similar Datasets

Project description:ObjectiveTo compare the reliability of different methods for measuring fat content of pancreas by MR modified Dixon(mDixon) Sequence and accurately evaluate pancreatic fat in as simple a way as possible.MethodsThis is a retrospective study, 64 patients were included in this study who underwent abdominal MR scan that contained the mDixon sequence from June 2019 to May 2020(Included 7 patients with type 2 diabetes and 4 patients with impaired glucose tolerance (IGT), they were admitted to hospital through the obesity clinic set up by endocrine department, all of them were initially diagnosed and untreated). All of the 64 patients were scanned in 3.0T MR (Philips Ingenia II) due to their condition, 10-34 slice pancreas images were obtained, which were different from each other. Three different methods of measurement were employed by two observers using Philips Intellispace Portal software: (1) All images (whole-pancreas) measurement, the whole-pancreatic fat fraction (wPFF) was calculated by software. (2) Interval slices measurement, that is half-pancreatic slices fat fraction (hPFF) measured in the same way, fat fraction obtained by the interlayer assay was calculated. (3) As usual, the fat content of pancreatic head, body and tail fat was measured respectively, and in order to improve credibility, we also measured head、 body and tail in every layer, and its average value was taken. The elapsed time of the above different measurement methods was recorded. Intra-group correlation coefficient (ICC) was used to analyze the consistency of the measured data within and between observers. T-tests and Friedman tests were applied to compare the difference of measured values among groups.ResultsNo matter in normal person or diabetic or IGT, hPFF has shown good stability (ICChPFF = 0.988), and there was no significant difference compared with wPFF. But the average fat percentage composition of head, body and tail were significantly different from wPFF and hPFF (P < 0.01). At the same time, compared with normal person, pancreatic fat content in IGT and diabetic patients showed progressive significance(P<0.05).ConclusionThe distribution of pancreatic fat is not uniform, the method of measuring half pancreas by interlayer data collection can reflect the fat content of the entire pancreas, this suggests that measuring 50% of the pancreas is sufficient, this method effectively saves time and effort without affecting the results, which may have a better clinical application prospect.

Project description:BackgroundHFrEF is a heterogenous condition with high mortality. We used serial assessments of 4210 circulating proteins to identify distinct novel protein-based HFrEF subphenotypes and to investigate underlying dynamic biological mechanisms. Herewith we aimed to gain pathophysiological insights and fuel opportunities for personalised treatment.MethodsIn 382 patients, we performed trimonthly blood sampling during a median follow-up of 2.1 [IQR:1.1-2.6] years. We selected all baseline samples and two samples closest to the primary endpoint (PEP; composite of cardiovascular mortality, HF hospitalization, LVAD implantation, and heart transplantation) or censoring, and applied an aptamer-based multiplex proteomic approach. Using unsupervised machine learning methods, we derived clusters from 4210 repeatedly measured proteomic biomarkers. Sets of proteins that drove cluster allocation were analysed via an enrichment analysis. Differences in clinical characteristics and PEP occurrence were evaluated.FindingsWe identified four subphenotypes with different protein profiles, prognosis and clinical characteristics, including age (median [IQR] for subphenotypes 1-4, respectively:70 [64, 76], 68 [60, 79], 57 [47, 65], 59 [56, 66]years), EF (30 [26, 36], 26 [20, 38], 26 [22, 32], 33 [28, 37]%), and chronic renal failure (45%, 65%, 36%, 37%). Subphenotype allocation was driven by subsets of proteins associated with various biological functions, such as oxidative stress, inflammation and extracellular matrix organisation. Clinical characteristics of the subphenotypes were aligned with these associations. Subphenotypes 2 and 3 had the worst prognosis compared to subphenotype 1 (adjHR (95%CI):3.43 (1.76-6.69), and 2.88 (1.37-6.03), respectively).InterpretationFour circulating-protein based subphenotypes are present in HFrEF, which are driven by varying combinations of protein subsets, and have different clinical characteristics and prognosis.Clinical trial registrationClinicalTrials.gov Identifier: NCT01851538https://clinicaltrials.gov/ct2/show/NCT01851538.FundingEU/EFPIA IMI2JU BigData@Heart grant n°116074, Jaap Schouten Foundation and Noordwest Academie.

Dataset Information

Repeatedly measured predictors: a comparison of methods for prediction modeling.

Background

Methods

Results

Conclusions

Publications

Repeatedly measured predictors: a comparison of methods for prediction modeling.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets