MATH 225N Week 4 Discussion: Probability
Required Resources
• Read/review the following resources for this activity:
• OpenStax Textbook: Chapter
... [Show More] 2, 3
Initial Post Instructions
Consider when you are listening to a new patient who says they have had a fever for 3 days. You would use your knowledge and experience with probability to come to a diagnosis. What changes those probabilities? Is it age, is it other symptoms, is it new study results? What is the source of these probabilities and what causes these probabilities to change? This discussion is not to debate diagnoses, but to focus on the sources and influences on probabilities.
NB: 2 Answers Displayed
Answer 1
A probabilistic approach to patient diagnosis is too often taken for granted by both physicians and patients. Although some physicians express their uncertainty as to the probability that the patient has a specified disease, using probability rather than ambiguous terms such as "probably" or "possibly" enables the clinician expresses uncertainty or certainty quantitatively (Liu et al., 2018). Considering the provided case in which the patient has had a fever for three days, the knowledge and experience with probability can be helpful in the diagnosis. Probabilities can be changed by how the patient describes the symptoms; for instance, a clear and accurate description of the symptoms leads to more certain probabilities being generated. Age can also be a factor given that some conditions become amplified with age, and new study results about the condition can also help inform the generation or more accurate probabilities.
The source of the probabilities would be clinical journals, besides one’s experience in the field. Tests conducted in a medical lab, besides other diagnostic methods like ultrasound, radiography, and magnetic resonance imaging can cause established probabilities to change. Clinical prediction rules related to the intersection of events and conditional probability can have a notable influence on probabilities (Sargolzaei et al., 2015; Holmes, Illowsky, & Dean, 2018).
References
Holmes, A., Illowsky, B. & Dean, S. (2018). Introductory business statistics. Houston, TX: OpenStax
Liu, S., Zeng, J., Gong, H., Yang, H., Zhai, J., Cao, Y., ... & Ding, X. (2018). Quantitative analysis of breast cancer diagnosis using a probabilistic modelling approach. Computers in biology and medicine, 92, 168-175.
Sargolzaei, S., Cabrerizo, M., Sargolzaei, A., Noei, S., Eddin, A. S., Rajaei, H., ... & Adjouadi, M. (2015). A probabilistic approach for pediatric epilepsy diagnosis using brain functional connectivity networks. BMC bioinformatics, 16(7), S9.
Answer 2
If I were considering a patient who presents with a fever for three days I would want to look further into diagnosing the cause of the fever using my experience. Keep in mind the values I am presenting are entirely fictional, and are used only to demonstrate the use of probabilities. Fever is commonly related to one of four body systems (emedicalhub, 2019) – respiratory, digestive, urogenital, or skin. In the following examples, let P(R) = % of patients with fever of respiratory etiology, P(D) = digestive etiology, P(U) = urogenital etiology, P(S) = skin etiology.
For example sake let’s say for all causes of fever that P(R) = 0.4, P(D) = 0.1, P(U) = 0.4, P(S) = 0.1. From this we know that there is an eighty percent chance that the cause of an unknown fever is from either a respiratory or urogenital etiology. I could say that P(R or U) = 0.8. Narrowing subsequent inquiry and testing to these two categories would increase my chance at successful diagnosis.
Now for example sake let’s say the patient also complains of painful urination. With this additional information that the patient has fever and dysuria, the probability of the fever being of urogenital origin might change to P(U) = 0.7. I could then look at the patient’s age and sex. Let’s say the patient presenting with fever and dysuria also happens to be an 87 year old female. My P(U) has now probably changed to P(U) = 0.9. If I know that this patient’s probability of having a symptomatic UTI resulting in fever is 90%, I could perform a UA and treat appropriately. These are just fictional values to present how probabilities in diagnosing change as more information is presented. Thank you, [Show Less]