Limits...
Doctor, what does my positive test mean? From Bayesian textbook tasks to personalized risk communication.

Navarrete G, Correia R, Sirota M, Juanchich M, Huepe D - Front Psychol (2015)

Bottom Line: This type of research aims to translate empirical findings into effective ways of providing risk information.In this paper, we concentrate on the communication of the result of medical tests and outline the epidemiological and test parameters that affect the predictive power of a test-whether it is correct or not.Building on this, we draw up recommendations for better practice to convey the results of medical tests that could inform health policy makers (What are the drawbacks of mass screenings?), be used by health practitioners and, in turn, help patients to make better and more informed decisions.

View Article: PubMed Central - PubMed

Affiliation: Psychology Department, Laboratory of Cognitive and Social Neuroscience, UDP-INECO Foundation Core on Neuroscience, Universidad Diego Portales Santiago, Chile.

ABSTRACT
Most of the research on Bayesian reasoning aims to answer theoretical questions about the extent to which people are able to update their beliefs according to Bayes' Theorem, about the evolutionary nature of Bayesian inference, or about the role of cognitive abilities in Bayesian inference. Few studies aim to answer practical, mainly health-related questions, such as, "What does it mean to have a positive test in a context of cancer screening?" or "What is the best way to communicate a medical test result so a patient will understand it?". This type of research aims to translate empirical findings into effective ways of providing risk information. In addition, the applied research often adopts the paradigms and methods of the theoretically-motivated research. But sometimes it works the other way around, and the theoretical research borrows the importance of the practical question in the medical context. The study of Bayesian reasoning is relevant to risk communication in that, to be as useful as possible, applied research should employ specifically tailored methods and contexts specific to the recipients of the risk information. In this paper, we concentrate on the communication of the result of medical tests and outline the epidemiological and test parameters that affect the predictive power of a test-whether it is correct or not. Building on this, we draw up recommendations for better practice to convey the results of medical tests that could inform health policy makers (What are the drawbacks of mass screenings?), be used by health practitioners and, in turn, help patients to make better and more informed decisions.

No MeSH data available.


Related in: MedlinePlus

Positive predictive value for three tests with a 100% sensitivity according to the rate of false positive (A) 0.1%, (B) 1%, and (C) 2%, and to the prevalence of the condition.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4585185&req=5

Figure 1: Positive predictive value for three tests with a 100% sensitivity according to the rate of false positive (A) 0.1%, (B) 1%, and (C) 2%, and to the prevalence of the condition.

Mentions: Of course, as often happens, if a medical test is not as reliable as the one used in the two examples above (100% sensitivity, and 0.1% false positive rate), a low positive predictive value appears even with common medical conditions. For example, see in Equation (6) the computation of the positive predictive value of a test aiming to detect a condition with a prevalence of 1 in 100, and a false positive rate as low as 1%. When the rate of false positives increases by 0.9%, the positive predictive value of the test decreases by 40%, dropping from 90 to 50%. In this context, a person receiving a positive test has only a 50% chance of actually having the condition.(6)p(H/D)=p(D&H)p(D)=11+0.99=0.5With all these examples, we are not implying that screening tests should not be trusted. We intend to outline the factors needed to be considered when using and interpreting medical test results. As we have seen, low prevalence rates, and their interaction with false positive rates, are generally guilty of decreasing the positive predictive value of a test: Figure 1 provides an illustration of this. The variability of positive predictive values of medical tests, according to the characteristics of the test and the prevalence of the condition, makes it hard for patients to decide whether to take the test and to assess their chances of having a condition when they test positive, particularly when the information given to them is generally too complicated to understand.


Doctor, what does my positive test mean? From Bayesian textbook tasks to personalized risk communication.

Navarrete G, Correia R, Sirota M, Juanchich M, Huepe D - Front Psychol (2015)

Positive predictive value for three tests with a 100% sensitivity according to the rate of false positive (A) 0.1%, (B) 1%, and (C) 2%, and to the prevalence of the condition.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4585185&req=5

Figure 1: Positive predictive value for three tests with a 100% sensitivity according to the rate of false positive (A) 0.1%, (B) 1%, and (C) 2%, and to the prevalence of the condition.
Mentions: Of course, as often happens, if a medical test is not as reliable as the one used in the two examples above (100% sensitivity, and 0.1% false positive rate), a low positive predictive value appears even with common medical conditions. For example, see in Equation (6) the computation of the positive predictive value of a test aiming to detect a condition with a prevalence of 1 in 100, and a false positive rate as low as 1%. When the rate of false positives increases by 0.9%, the positive predictive value of the test decreases by 40%, dropping from 90 to 50%. In this context, a person receiving a positive test has only a 50% chance of actually having the condition.(6)p(H/D)=p(D&H)p(D)=11+0.99=0.5With all these examples, we are not implying that screening tests should not be trusted. We intend to outline the factors needed to be considered when using and interpreting medical test results. As we have seen, low prevalence rates, and their interaction with false positive rates, are generally guilty of decreasing the positive predictive value of a test: Figure 1 provides an illustration of this. The variability of positive predictive values of medical tests, according to the characteristics of the test and the prevalence of the condition, makes it hard for patients to decide whether to take the test and to assess their chances of having a condition when they test positive, particularly when the information given to them is generally too complicated to understand.

Bottom Line: This type of research aims to translate empirical findings into effective ways of providing risk information.In this paper, we concentrate on the communication of the result of medical tests and outline the epidemiological and test parameters that affect the predictive power of a test-whether it is correct or not.Building on this, we draw up recommendations for better practice to convey the results of medical tests that could inform health policy makers (What are the drawbacks of mass screenings?), be used by health practitioners and, in turn, help patients to make better and more informed decisions.

View Article: PubMed Central - PubMed

Affiliation: Psychology Department, Laboratory of Cognitive and Social Neuroscience, UDP-INECO Foundation Core on Neuroscience, Universidad Diego Portales Santiago, Chile.

ABSTRACT
Most of the research on Bayesian reasoning aims to answer theoretical questions about the extent to which people are able to update their beliefs according to Bayes' Theorem, about the evolutionary nature of Bayesian inference, or about the role of cognitive abilities in Bayesian inference. Few studies aim to answer practical, mainly health-related questions, such as, "What does it mean to have a positive test in a context of cancer screening?" or "What is the best way to communicate a medical test result so a patient will understand it?". This type of research aims to translate empirical findings into effective ways of providing risk information. In addition, the applied research often adopts the paradigms and methods of the theoretically-motivated research. But sometimes it works the other way around, and the theoretical research borrows the importance of the practical question in the medical context. The study of Bayesian reasoning is relevant to risk communication in that, to be as useful as possible, applied research should employ specifically tailored methods and contexts specific to the recipients of the risk information. In this paper, we concentrate on the communication of the result of medical tests and outline the epidemiological and test parameters that affect the predictive power of a test-whether it is correct or not. Building on this, we draw up recommendations for better practice to convey the results of medical tests that could inform health policy makers (What are the drawbacks of mass screenings?), be used by health practitioners and, in turn, help patients to make better and more informed decisions.

No MeSH data available.


Related in: MedlinePlus