Limited information is available about predictors of short-term outcomes in patients with exacerbation of chronic obstructive pulmonary disease (eCOPD) attending an emergency department (ED). Such information could help stratify these patients and guide medical decision-making. The aim of this study was to develop a clinical prediction rule for short-term mortality during hospital admission or within a week after the index ED visit.

Variable selection is an important step to end up with good prediction models. LASSO regression models are one of the most commonly used methods for this purpose, for which cross‐validation is the most widely applied validation technique to choose the tuning parameter . Validation techniques in a complex survey framework are closely related to “replicate weights”. However, to our knowledge, they have never been used in a LASSO regression context. Applying LASSO regression models to complex survey data could be challenging. The goal of this paper is twofold. On the one hand, we analyze the performance of replicate weights methods to select the tuning parameter for fitting LASSO regression models to complex survey data. On the other hand, we propose new replicate weights methods for the same purpose. In particular, we propose a new design‐based cross‐validation method as a combination of the traditional cross‐validation and replicate weights. The performance of all these methods has been analyzed and compared by means of an extensive simulation study to the traditional cross‐validation technique to select the tuning parameter for LASSO regression models. The results suggest a considerable improvement when the new proposal design‐based cross‐validation is used instead of the traditional cross‐validation.

In medical practice many, essentially continuous, clinical parameters tend to be categorised by physicians for ease of decision-making. Indeed, categorisation is a common practice both in medical research and in the development of clinical prediction rules, particularly where the ensuing models are to be applied in daily clinical practice to support clinicians in the decision-making process. Since the number of categories into which a continuous predictor must be categorised depends partly on the relationship between the predictor and the outcome, the need for more than two categories must be borne in mind. We propose a categorisation methodology for clinical-prediction models, using Generalised Additive Models (GAMs) with P-spline smoothers to determine the relationship between the continuous predictor and the outcome. The proposed method consists of creating at least one average-risk category along with high- and low-risk categories based on the GAM smooth function. We applied this methodology to a prospective cohort of patients with exacerbated chronic obstructive pulmonary disease. The predictors selected were respiratory rate and partial pressure of carbon dioxide in the blood (PCO2), and the response variable was poor evolution. An additive logistic regression model was used to show the relationship between the covariates and the dichotomous response variable. The proposed categorisation was compared to the continuous predictor as the best option, using the AIC and AUC evaluation parameters. The sample was divided into a derivation (60%) and validation (40%) samples. The first was used to obtain the cut points while the second was used to validate the proposed methodology. The three-category proposal for the respiratory rate was ≤ 20;(20,24];> 24, for which the following values were obtained: AIC=314.5 and AUC=0.638. The respective values for the continuous predictor were AIC=317.1 and AUC=0.634, with no statistically significant differences being found between the two AUCs (p =0.079). The four-category proposal for PCO2 was ≤ 43;(43,52];(52,65];> 65, for which the following values were obtained: AIC=258.1 and AUC=0.81. No statistically significant differences were found between the AUC of the four-category option and that of the continuous predictor, which yielded an AIC of 250.3 and an AUC of 0.825 (p =0.115). Our proposed method provides clinicians with the number and location of cut points for categorising variables, and performs as successfully as the original continuous predictor when it comes to developing clinical prediction rules.

Los modelos predictivos permiten estratificar a la población en función de los cuidados sanitarios que requerirán el año siguiente. Ofrecen a las organizaciones sanitarias la oportunidad de actuar proactivamente, diseñando intervenciones específicas, adecuadas al nivel de necesidad de los distintos grupos de personas. La Estrategia de Cronicidad de Euskadi plantea el empleo de estos modelos, integrándolos con otras políticas y áreas de cambio. En 2010 se realizó por primera vez la categorización prospectiva de toda la población adscrita a Osakidetza, por medio del sistema Johns Hopkins Adjusted Clinical Groups Predictive Model (ACG-PM). Para ello se empleó, de forma secundaria, información ya registrada, procedente de las historias clínicas informatizadas de atención primaria y de altas de hospital. Este trabajo plantea las ventajas de la utilización combinada de diferentes fuentes de información y describe la aplicación de la estratificación poblacional en 3 programas, dirigidos a enfermos crónicos con diferente carga de comorbilidad.