Clinical decision rules in radiology

Clinical decision rules in primary care: necessary investments for sustainable healthcare

Clinical judgement in primary care is more often decisive than in the hospital. Clinical decision rules (CDRs) can help general practitioners facilitating the work-through of differentials that follows an initial suspicion, resulting in a concrete 'course of action': a 'rule-out' without further testing, a need for further testing, or a specific treatment. However, in daily primary care, the use of CDRs is limited to only a few isolated rules. In this paper, we aimed to provide insight into the laborious path required to implement a viable CDR. At the same time, we noted that the limited use of CDRs in primary care cannot be explained by implementation barriers alone. Through the case study of the Oudega rule for the exclusion of deep vein thrombosis, we concluded that primary care CDRs come out best if they are tailor-made, taking into consideration the specific context of primary health care. Current CDRs should be evaluated frequently, and future decision rules should anticipate the latest developments such as the use of point-of-care (POC) tests. Hence, such new powerful diagnostic CDRs could improve and expand the possibilities for patient-oriented primary care.

Machine Learning Principles for Radiology Investigators

Artificial intelligence and deep learning are areas of high interest for radiology investigators at present. However, the field of machine learning encompasses multiple statistics-based techniques useful for investigators, which may be complementary to deep learning approaches. After a refresher in basic statistical concepts, relevant considerations for machine learning practitioners are reviewed: regression, classification, decision boundaries, and bias-variance tradeoff. Regularization, ground truth, and populations are discussed along with compute and data management principles. Advanced statistical machine learning techniques including bootstrapping, bagging, boosting, decision trees, random forest, XGboost, and support vector machines are reviewed along with relevant examples from the radiology literature.

An online evidence-based decision support system for distinguishing benign from malignant vertebral compression fractures by magnetic resonance imaging feature analysis

Decision support systems have been used to promote the practice of evidence-based medicine. Computer-assisted diagnosis can serve as one element of evidence-based radiology. One area where such tools may provide benefit is analysis of vertebral compression fractures (VCFs), which can be a challenge in MRI interpretation. VCFs may be benign or malignant in etiology, and several MRI features may help to make this important distinction. We describe a web-based decision support system for discriminating benign from malignant VCFs as a prototype for a more general diagnostic decision support framework for radiologists. The system has three components: a feature checklist with an image gallery derived from proven reference cases, a prediction model, and a reporting mechanism. The website allows users to input the findings for a case to be interpreted using a structured feature checklist. The image gallery complements the checklist, for clarity and training purposes. The input from the checklist is then used to calculate the likelihood of malignancy by a logistic regression prediction model. Standardized report text is generated that summarizes pertinent positive and negative findings. This computer-assisted diagnosis system demonstrates the integration of three areas where diagnostic decision support can aid radiologists: first, in image interpretation, through feature checklists and illustrative image galleries; second, in feature-based prediction modeling; and third, in structured reporting. We present a diagnostic decision support tool that provides radiologists with evidence-based guidance for discriminating benign from malignant VCF. This model may be useful in other difficult-diagnosis situations and requires further clinical testing.

[Evidence-based radiology for diagnostic imaging: what it is and how to practice it]

Evidence-based radiology is defined as the decision that results from integrating clinical information to select the most appropriate imaging test on the basis of the best available evidence, the physician's experience, and the patient's expectations. The practice of evidence-based radiology consists of five steps: formulating the question, performing an efficient search of the literature, critically evaluating the literature, applying the results of the search and evaluation while taking into account our experience and the patient's values, and evaluating the results obtained within our own practice. In diagnostic imaging, the number of resources available for evidence-based radiology is increasing: apart from books, articles, and web pages on this subject, evidence-based radiology is receiving more attention at diagnostic imaging conferences. The principles of evidence-based radiology will help promote the appropriate use of resources, greatly benefiting patients (decreasing the use of examinations that use ionizing radiation), professionals (less overload), and managers (more efficient use of resources).

Sensitivity and specificity: imperfect predictors of guideline utility in radiology

OBJECTIVE

Many "guideline-development studies" have presented the sensitivity (SN) and specificity (SP) of a new decision tool to describe the potential improvements in utilisation of imaging techniques as a result of adopting the new guideline. However, SN and SP are measures designed to assess how well a new guideline compares with a gold standard. These measures do not evaluate how many patients with a positive test actually have the disease; nor do they evaluate how many patients with a negative test do not have the disease. To evaluate these characteristics of a decision tool, other measures, namely the positive predictive value (PPV) and negative predictive value (NPV), should be calculated. This report highlights some of the main methodological challenges in interpretation of the studies that attempt to evaluate the development of an imaging guideline and the effectiveness of an imaging guideline in real world practice.

METHODS

We define four key measures of a decision tool: SN, SP, PPV and NPV. Using data from two hypothetical populations, we explain how these measures can be calculated and interpreted. We place special emphasis on the purpose of and differences between the SN-SP and PPV-NPV.

RESULTS

Borrowing information from two studies, we demonstrate how these measures should be used in the radiology healthcare services research to evaluate decision guidelines.

CONCLUSION

The use of appropriate measures for the specific question at hand will ensure the guidelines are useful, safe, cost reducing and effective clinical tools.

Too much of a good thing is wonderful? A conceptual analysis of excessive examinations and diagnostic futility in diagnostic radiology

It has been argued extensively that diagnostic services are a general good, but that it is offered in excess. So what is the problem? Is not "too much of a good thing wonderful", to paraphrase Mae West? This article explores such a possibility in the field of radiological services where it is argued that more than 40% of the examinations are excessive. The question of whether radiological examinations are excessive cries for a definition of diagnostic futility. However, no such definition is found in the literature. As a response, this article addresses the issue of diagnostic futility in five steps. First, it investigates whether the concept of therapeutic futility can be adapted to diagnostics. A closer analysis of the concept of therapeutic futility reveals that this will not do the trick. Second, the article scrutinizes whether there are sources for clarifying diagnostic futility in the extensive debate on excessive radiological examination. Investigating the debate's terms and definitions reveals a disparate terminology and no clear concepts. On the contrary, the study uncovers that quite different and incompatible issues are at stake. Third, the article examines a procedural approach, which is widely used for settling controversies over utility by focusing on the role of the professionals. On scrutiny however, a procedural approach will not solve the problem in diagnostics. Fourth, a value analysis reveals how we have to decide on the negative value of excessive examinations before we can measure excess. The final and constructive part presents a definition of diagnostic futility drawing upon the lessons from the previous analytical steps. Altogether, too much radiological examination is not a good thing. This is simply because radiological examinations are not unanimously good. Excessive radiological examinations can be defined, but not by one simple general and value-neutral definition. We have to settle with contextually framed value-related definitions. Such definitions will state how bad "too much of a good thing" is and make it possible to assess how much of the bad thing there is. Hence we have to know how bad it is before we can tell how much of it there is in the world.