Fundamentals of Clinical Prediction Modeling for the Neurosurgeon

Abstract

Clinical prediction models in neurosurgery are increasingly reported. These models aim to provide an evidence-based approach to the estimation of the probability of a neurosurgical outcome by combining 2 or more prognostic variables. Model development and model reporting are often suboptimal. A basic understanding of the methodology of clinical prediction modeling is needed when interpreting these models. We address basic statistical background, 7 modeling steps, and requirements of these models such that they may fulfill their potential for major impact for our daily clinical practice and for future scientific work.

Application of clinical prediction modeling in pediatric neurosurgery: a case study

“…For the development of prediction models with regression analysis, it is advised to systematically follow distinct steps (Fig. 1) [5]. However, prior to the start of the development of a new prediction model, existing models should be searched for [5].…”

“…1) [5]. However, prior to the start of the development of a new prediction model, existing models should be searched for [5]. It is recommended to validate and/or update existing models instead of developing de novo prediction models.…”

“…The objective of the model should be clearly defined at the start, e.g., what is the clinical decision the model needs to support. Next, a description on how patients were selected for inclusion in the prediction analysis is imperative, because data are often primarily collected for other purposes and different study designs have different consequences on the interpretation of the results [5]. For example, for a diagnostic prediction model, a cross-sectional cohort study design is preferred, whereas for a prognostic prediction model, a longitudinal prospective cohort study design is preferred.…”

“…Many statistical approaches can be used to develop a prediction model, including but not limited to regression analysis [5]. Detailed statistical output is often presented in the manuscript or in its appendix which is needed to make an adequate evaluation of the presented prediction model.…”

Clinical prediction models

“…Published work does not always provide sufficient details of the model, although this is necessary to judge the quality of the prediction model. More details of the presented methodological concerns and others including the selection of candidate prognostic variables and evaluating model performance measures have been recently published and illustrated with clinical examples specifically for neurosurgeons [1]. Furthermore, 7 key steps for the development and evaluation of a neurosurgical clinical prediction model are tabulated for a quick overview.…”