Abstract
We introduce the basic analytical tools used in medical decision making such as the a priori probability of illness and the sensitivity and specificity of a test. Each chapter provides illustrations, examples and exercises to sharpen the reader’s understanding of the issues discussed.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Dooling, R.—White Man’s Grave—Picador, New York—1995, p. 70.
- 2.
\( 30\cdot 0.8+10\cdot 0.2=26 \)
- 3.
\( {\left(30-26\right)}^2\cdot 0.8+{\left(10-26\right)}^2\cdot 0.2=64 \)
- 4.
AWMF = ‘Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften’, Association of the Scientific Medical Societies in Germany
References
AWMF (Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften). (2002). PSA-Bestimmung in der Prostatakarzinomdiagnostik. Leitlinie-Nr. 043/036.
Eddy, D. M. (1982). Probabilistic reasoning in clinical medicine: Problems and opportunities. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgment under uncertainty: Heuristics and biases (pp. 249–267). Cambridge: Cambridge University Press.
Fletcher, R. H., Fletcher, S. W., & Wagner, E. H. (1996). Clinical epidemiology: The essentials. Baltimore, MD: Williams & Wilkins.
Gigerenzer, G., & Hoffrage, U. (1995). How to improve Bayesian reasoning without instruction: Frequency formats. Psychological Review, 102(4), 684–704.
Hoffman, R. M., Gilliland, F. D., Adams-Cameron, M., Hunt, W. C., & Key, C. R. (2002). Prostate-specific antigen testing accuracy in community practice. BMC Family Practice, 3(19), 1–8.
Hunink, M., Glasziou, P., Siegel, J., Weeks, J., Pliskin, J., Elstein, A., et al. (2004). Decision making in health and medicine. Cambridge: Cambridge University Press.
Kling, A. (2010). I thought you might be interested in Doctors’ Statistical Ignorance. EconLog. Library of Economics and Liberty. http://econlog.econlib.org/archives/2007/10/doctors_statist.html. Accessed 31 August 2010
Robert Koch Institut. (2010). Verbreitung von Krebserkrankungen in Deutschland. Entwicklung der Prävalenzen zwischen 1990 und 2010. Beiträge zur Gesundheitsberichterstattung des Bundes. Berlin: RKI.
Sackett, D. L., Guyatt, G. H., Haynes, R. B., & Tugwell, P. (1991). Clinical epidemiology: A basic science for clinical medicine. Philadelphia: Lippincott Williams & Wilkins.
Sox, H. S., Blatt, M. A., Higgins, M. C., & Marton, K. I. (2007). Medical decision making. Philadelphia: American College of Physicians, Butterworth-Heinemann (Reprint, originally published: 1988).
Van Gossum, A., Munoz-Navas, M., Fernandez-Urien, I., et al. (2009). Capsule endoscopy versus colonoscopy for the detection of polyps and cancer. The New England Journal of Medicine, 361(3), 264–270.
Weinstein, M. C., Fineberg, H. V., Elstein, A. S., Frazier, H. S., Neuhauser, D., Neutra, R. R., et al. (1980). Clinical decision analysis. Philadelphia: W.B. Saunders.
Author information
Authors and Affiliations
Exercises
Exercises
-
1.
There is a diagnostic test for prostate cancer which measures the concentration of PSA (prostate-specific antigens) in the blood. If the concentration exceeds a certain threshold, the physician applies advanced diagnostics and performs a biopsy. A positive outcome of the biopsy leads to a radical prostatectomy. If the PSA concentration is unremarkable, the physician will take another blood test one year later.
To determine the quality of the PSA test, Hoffman et al. (2002) conducted the test as well as a biopsy on 2620 men of age 40 or older in the U.S. state of New Mexico. The biopsies revealed prostate cancer in 930 men. With a cutoff value of 4Â ng/ml, the PSA test detected 800 true cases of sick men and 558 true cases of healthy men.
-
(a)
Calculate sensitivity and specificity for a critical serum concentration of 4Â ng/ml.
-
(b)
Fill out the 2 × 2 table for four age groups if the following 10-year prevalence rates apply for Germany:
Age group
Prevalence rate (in percent)
<60
0.1
60–69
2.3
70–79
4.9
>80
5.5
All
0.8
-
(c)
Use the positive likelihood ratio to calculate the odds of a true positive case for each age group. Explain the relationship between the a priori and the a posteriori odds of identifying a true positive patient.
-
(d)
Regarding the calculated a posteriori odds for the four age groups, how useful is the test result given a cutoff value of 4Â ng/ml? Which parameters would improve the test?
-
(e)
Calculate a prevalence rate at which 90 out of 100 men who were tested positive will in fact be sick (at a cutoff value of 4Â ng/ml). Compared to the pertinent prevalence rate in Germany, what conclusions can be drawn from this case about the usefulness of the PSA test as a screening method?
-
(f)
Early AWMFFootnote 4 (2002) guidelines advised a critical value of 4Â ng/ml for all age groups. Argue why this is not tenable, given the different prevalence rates in the age groups.
-
(g)
According to these guidelines, family background (genetics) and fat intake are further risk factors for prostate cancer. Is it justified to offer the PSA test to targeted subgroups of the population?
-
(a)
-
2.
The following table lists the sensitivity and specificity for PSA cutoff values between 1 and 20Â ng/ml (see Hoffman et al. 2002) for men of age 40 and older.
PSA
Sensitivity
Specificity
1
98
9
2
95
20
3
91
26
4
86
33
5
75
44
6
63
57
7
56
66
8
49
74
9
42
80
10
38
84
15
23
93
20
17
97
-
(a)
Draw the ROC curve.
-
(b)
Calculate the positive likelihood ratio for each PSA cutoff value. How do sensitivity and specificity change along the ROC curve?
-
(c)
Which consequences would arise from reducing the cutoff value from 4 to 2Â ng/ml?
-
(d)
In 2003, the German parliament called for coverage of PSA tests under the Social Health Insurance for all men older than 45. State the pros and cons of such a policy.
-
(a)
-
3.
Prove that a test’s likelihood ratio improves if either sensitivity or specificity increases.
-
4.
Describe the distribution of the diagnostic measurement values for the sick and the healthy in Fig. 2.1 for a perfect test (sensitivity and specificity equal to unity). Where is the optimal cutoff value for testing?
Rights and permissions
Copyright information
© 2017 Springer-Verlag GmbH Germany
About this chapter
Cite this chapter
Felder, S., Mayrhofer, T. (2017). Basic Tools in Medical Decision Making. In: Medical Decision Making. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-53432-8_2
Download citation
DOI: https://doi.org/10.1007/978-3-662-53432-8_2
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-53431-1
Online ISBN: 978-3-662-53432-8
eBook Packages: Economics and FinanceEconomics and Finance (R0)