Cancer patients and their physicians face many decisions related to treatment options. Depending on the type of cancer, the location of the cancer, and patient history, treatment options can include surgery, chemotherapy, radiation therapy, and immunotherapy. The probability of the success of these treatments is affected by many factors based on the type of cancer and overall health of the patient. Medical decision makers often measure the expected effectiveness of different types of treatments using a metric called quality adjusted life year (QALY). One QALY is equivalent to one year of life in perfect health. Decision makers use QALYs to approximate the possible outcomes on the patient from different treatment options.

Suppose that a physician is working with a cancer patient to determine the best course of treatment. The physician and the patient need to decide whether to have an exploratory surgery procedure to determine the spread of the cancer or to follow an active surveillance strategy in which no surgery is performed but the patient has regular tests to track progression of the cancer. If exploratory surgery is chosen, the results of the surgery can be favorable (F), unfavorable (U), or show that no further treatment is needed (N). Past data suggests that P(F) = 0.5, P(U) = 0.2, and P(N) = 0.3. If the exploratory surgery is favorable or unfavorable, then one of two treatments will be used, Treatment A (A) or Treatment B (B) using the following conditional probabilities:

Possible patient outcomes are measured using QALYs, and this depends on both the result of the exploratory surgery and the treatment option required as shown in the following table.

If active surveillance is chosen, then the cancer can either progress (P) or not progress (NP) where P(P) = 0.4 and P(NP) = 0.6. If the cancer does not progress, then no fur- ther treatment is needed. If the cancer does progress, then either Treatment A or Treat- ment B will be required where P(AIP) = 0.7 and P(BIP) = 0.3. The possible resulting QALYs for this decision alternative follow. 

a. Construct a decision tree for this problem.

b. Using the expected value approach and QALYs as the outcome measure, would you recommend exploratory surgery or active surveillance? What is the expected QALYs associated with this decision?

Transcribed Image Text:

P(A|F) = 0.4 P(BIF) = 0.6 P(AIU) = 0.8 P(BIU) = 0.2