The Proliferation of Reports on Clinical Scoring Systems

Abstract

An ongoing trend in medicine toward evidence-based practice emphasizes efficiency and algorithmic approaches toward diagnosis and treatment. This approach has a number of potential benefits including cost-savings, standardization of practice, and improvement in quality of care. 1 A proliferation of clinical scoring systems has in part been a consequence of this movement. These scoring systems are designed to inform clinical decision-making and may involve tools for diagnosis (eg, Wells Criteria), prognosis of disease (eg, the Model for End-Stage Liver Disease score), and future risk of developing a disease (eg, Atherosclerotic Cardiovascular Disease risk score).Clinical scoring systems vary from those that are widely recognized and used (eg, National Institutes of Health Stroke Scale, APGAR [Appearance, Pulse, Grimace, Activity, Respiration]) to those that are highly specialized and used less commonly (eg, A 2 DS 2 [age, atrial fibrillation, dysphagia, sex, and severity]). However, there has been a large recent increase in scoring systems without evidence of any clear clinical benefit for most. Each week, new scoring systems are published in the medical literature. In this Viewpoint, we assess this phenomenon and explore the clinical usefulness of these scores.A search of the MEDLINE/PubMed database using various search terms (eAppendix in the Supplement) to identify peer-reviewed articles related to clinical scoring systems resulted in approximately 250 000 publications since 1965, with an almost exponential increase in yearly publications of scoring systems over time (Figure). Until the mid-1990s, fewer than 5 publications per 1000 total publications were identified, whereas in 2018, more than 20 publications per 1000 publications were identified.

Application of human factors to improve usability of clinical decision support for diagnostic decision-making: a scenario-based simulation study

“…In seven cases, study participants chose no further testing, whereas PE-Dx recommended a D-dimer as the Wells’ score was below two with PERC positive; in one case, the physician ordered a CTA, whereas PE-Dx suggested a D-dimer. It is possible that in those eight cases (out of 160 cases) physicians disagree with the recommendation from the CDS or did not trust the CDS 62. But we are unsure why study participants made those decisions as we did not ask them to think aloud so that we could obtain accurate estimates of time needed to complete tasks.…”

System for High-Intensity Evaluation During Radiation Therapy (SHIELD-RT): A Prospective Randomized Study of Machine Learning–Directed Clinical Evaluations During Radiation and Chemoradiation

“…Randomized interventional data are limited to a recently published study applying ML to intraoperative hypotension prediction. 18 High quality data are necessary to demonstrate the prospective accuracy and value of the expanding number of published ML algorithms, particularly given the well-documented limitations of retrospective validation 17,23,24 . Despite their general proliferation, the value of AI and ML algorithms in medicine remains generally unverified, and their deployment remains accordingly limited.…”

“…9,10 Despite a growing body of retrospective medical ML literature, 11,12 prospective evaluation remains tremendously limited, primarily to diagnostic fields. [13][14][15][16] Despite the need for prospective interventional trials, 17 there have been few, including a recent trial effectively predicting intraoperative hypotension utilizing arterial catheter sensor data. 18 There are even fewer studies investigating the use of ML to improve clinical outcomes, which may be related to challenges in clinical implementation and viability of systematic management strategies.…”

CHADS2 and CHA2DS2-VASc scores as tools for long-term mortality prognosis in patients with typical atrial flutter after catheter ablation

“…Similarly, the multivariable Cox proportional hazard analysis showed an independent predictive value of 3 components of the CHADS 2 score (C, 422 401 345 288 242 185 140 102 70 39 13 1 0 16 14 13 13 10 6 2 1 0 1 P = 0.75 Survival probability B a practical point of view, new prognostic scores tend to have limited popularity, which further limits their usefulness. 18 In contrast, CHADS 2 is already widely known and straightforward, both in out -patient and in -hospital settings as the required data is readily available. This makes our results more pertinent to clinical practice.…”