Variable selection methods were poorly reported but rarely misused in major medical journals: Literature review.

Objective This work presents a review of the literature on reporting, practice and misuse of knowledge-based and data-driven variable selection methods, in five highly cited medical journals, considering recoding and interaction unlike previous reviews. Study Design and Setting Original observational studies with a predictive or explicative research question with multivariable analyses published in N. Engl. J. Med., Lancet, JAMA, Br. Med. J. and Ann. Intern. Med. between 2017 and 2019 were searched. Article screening was performed by a single reader, data extraction was performed by two readers and a third reader participated in case of disagreement. The use of data-driven variable selection methods in causal explicative questions was considered as misuse. Results 488 articles were included. The variable selection method was unclear in 234 (48%) articles, data-driven in 78 (16%) articles and knowledge-based in 176 (36%) articles. The most common data-driven methods were: Univariate selection (n = 22, 4.5%) and model comparisons or testing for interaction (n = 17, 3.5%). Data-driven methods were misused in 51 (10.5%) of articles. Conclusion Overall reporting of variable selection methods is insufficient. Data-driven methods seem to be used only in a minority of articles of the big five medical journals.

A prospective study of current diagnostic procedures for assessment of hepatic metastases in colorectal cancers.

In a prospective study of 305 patients with colorectal cancers, we assessed the diagnostic value of ultrasonography and laboratory tests. In each case laparotomy was carried out and the presence of liver metastases was established in 47 patients. The results show that the laboratory tests alone are not sufficiently accurate to detect liver metastases. Additional accuracy can be obtained by the combined use of a single liver imaging test (echography) and selected laboratory tests (C.E.A., gamma GT, Alkaline Phosphatase).

[Comparative study of ultrasonography and scintigraphy in liver metastases detection in cases of colorectal carcinoma (author's transl)]. / Etude comparative de l'échographie et de la scintigraphie dans le dépistage des métastases hépatiques.

A comparative study has been realized to test the accuracy of ultrasonography and scintigraphy for detecting the presence of liver metastases in 305 patients with colorectal carcinomas. Presence or absence of hepatic metastases has been affirmed by laparotomy and biopsy. In the 47 cases with metastases, the sensitivity of ultrasonography was 93%, that of scintigraphy being 76%. In 258 cases without metastases the specificity of ultrasonography was 97%, that of scintigraphy being 92%. These results, confirmed by report in the published literature, show that ultrasonography should be the first examination for suspected hepatic metastases.

Natural history of colorectal carcinoma with untreated liver metastases.

In a series of 474 patients operated on for colonic and rectal carcinoma, 65 were shown to have hepatic metastases. Factors influencing spontaneous survival were evaluated in 56 patients. This study emphasizes the dominant role of the amount of liver involvement. A clinical staging referring to the size and number of liver deposits, to the presence or absence of jaundice and/or malignant ascites, and to liver function tests abnormalities provides an available classification which can be closely related to the outcome of such patients. Thus, it yields a valuable basis to avoid misleading prognostic data and to substantiate the effectiveness of aggressive treatments in case of carefully selected patients.