A simple, nonparametric discrimination procedure ,was developed and tested for use in discriminating between two populations, especially in those cases where the properties of the distributions lead to unsatisfactory results by classical discrimination procedures.The procedure is based on transferring the conventional concept of a decision limit from one to several variables, with combinations (in set theory terms: unions and intersections) of multidimensional intervals serving äs the discriminant regions (referred to here äs simple discriminant regions).The procedure was applied to two groups of patients, alcoholics and non-alcoholics. A level of efficiency 20% higher than that attainable with conventional procedures was obtained using only six clinical chemical parameters.Hansert, Federkiel and Stamm: A new procedure for discriminating between two patient populations Angewendet wurde das Verfahren auf je eine Stichprobe von Patienten, die als Alkoholiker bzw. von Patienten, die als Nichtalkoholiker eingestuft waren. Erreicht wurde mit dem einfachen Verfahren eine um etwa 20% höhere Effizienz als bei herkömmlichen Verfahren, und zwar auf der Bssis von nur 6 klinischchemischen Kenngrößen.Das Verfahren ist außerdem viel flexibler als herkömmliche Verfahren, da es ohne prinzipielle Schwierigkeiten gestattet, Optimierungsaufgabeh für jedes der drei gebräuchlichen Diskriminationskriterien (Seüsitivität, Spezifität und Effizienz) zu lösen, wobei noch freie Auswahl der Bereiche für die Entscheidungsgrenze und für die Ordnung des Diskriminanzbereiches möglich ist.Bei der Anwendung zeigten sich starke Überschneidungen der gegebenen zwei Stichproben in allen zugrundegelegten Dimensionen. So können z.B. bei Vorgabe von 100% Sensitivität bzw. 100% Spezifität mit den 6 wichtigsten Kenngrößen maximal etwa 53% Spezifität bzw. 56% Sensitivität erreicht werden.Diskutiert werden außerdem die Analogie des Diskriminationsprinzips zum Prinzip eines statistischen Tests (das Verhältnis von Ausschluß und Erkennung des Alkoholismus). Contents 1.
The knowledge of factors determining the risk of postoperative myocardial failure (MF) should allow a more rational approach to the timing and the management of mitral valve replacement (MVR). Using multivariate logistic regression analysis the influence of 41 preoperative and perioperative variables on MF was assessed in a training group of 353 consecutive patients undergoing isolated primary MVR between 6/76 and 12/82. Early MF mortality was 4.2%. Strongest independent preoperative predictors of MF were advanced NYHA functional class (p less than 0.001), hepatomegaly (p = 0.001), and reduced body weight (p = 0.01). Amongst preoperative and perioperative variables independent determinants of MF were NYHA functional class (p less than 0.001), hepatomegaly (p = 0.002), hypotension during extracorporeal circulation (ECC) (p = 0.005), body weight (p = 0.007), ECC duration (p = 0.008), female sex (p = 0.061) and the absence of cardioplegia (p = 0.065). From the combination of these determinants estimates of the probability of MF were calculated and adjoined to low or high risk by means of an optimum cutoff point. The sensitivity of this test performed before and after operation was 0.80 and 0.93, the specificity 0.92 and 0.94, respectively. The reliability of this prognostic test was prospectively evaluated on data of 107 consecutive MVR patients between 1/83 and 12/84. The observed diagnostic characteristics of the test group were comparable to those predicted from the training group. Multivariate logistic regression analysis selects independent determinants, estimates the risk of MF or other modes of postoperative events and identifies patients with low or high risk with a definable validity as an objective aid for medical decision-making.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.