Sample Size Considerations for Diagnostic Tests
Innovations in diagnostic techniques are increasing worldwide. Sensitivity and specificity are used to measure the accuracy of new dichotomous outcome diagnostic tests in the presence of an existing gold standard. The first question to be dealt is what nu
- PDF / 272,257 Bytes
- 24 Pages / 439.37 x 666.142 pts Page_size
- 91 Downloads / 269 Views
Sample Size Considerations for Diagnostic Tests Application to Sensitivity and Specificity Rajeev Kumar Malhotra
Abstract Innovations in diagnostic techniques are increasing worldwide. Sensitivity and specificity are used to measure the accuracy of new dichotomous outcome diagnostic tests in the presence of an existing gold standard. The first question to be dealt is what number of subjects is sufficient to attain adequate power in the case of hypothesis testing. This chapter explains sample size issues for estimating the sensitivity and specificity and compares sensitivity and specificity under different goals for two commonly used study designs, case–control and prospective designs in paired and unpaired subjects. The chapter also explains and compares the three methods to control uncertainty under the prospective design.
Introduction New innovations in diagnostic techniques are increasing due to advancements in technology. Researchers develop newer diagnostic techniques as surrogates for existing gold standards that are difficult to adopt in practice, expensive, not widely available, invasive, risky, and time consuming. For example, the diagnosis of pancreatic carcinoma can be confirmed only by invasive methods such as laparotomy or at autopsy. New diagnostic screening tests for detecting disease should be tested against existing reference diagnostic methods or with an established gold standard. The gold standard may be a more expensive diagnostic method or a combination of tests (combined reference standard) or may be available from clinical follow-up, surgical verification, biopsy, autopsy or a panel of experts. As we have seen in previous chapters, sensitivity and specificity are two indices that measure the accuracy of dichotomous outcome diagnostic tests with respect to an
R.K. Malhotra (*) Department of Biostatistics and Medical Informatics, University College of Medical Sciences, New Delhi, India e-mail: [email protected] S.A.R. Doi and G.M. Williams (eds.), Methods of Clinical Epidemiology, Springer Series on Epidemiology and Public Health, DOI 10.1007/978-3-642-37131-8_7, © Springer-Verlag Berlin Heidelberg 2013
79
80
R.K. Malhotra
existing gold standard. The first question to be dealt with is what is an adequate number of subjects needed for valid estimation of the diagnostic indices (sensitivity or specificity) or to attain adequate power in case of hypothesis testing. This chapter explains the sample size for sensitivity and specificity based on various goals and study designs. Sample size estimation is important not only for adequate power but also for other estimates such as costs and duration of the study. Arbitrary sample size based on convenience fails to achieve adequate power. Small sample sizes produces imprecise estimates and an unduly large sample is a waste of resources especially when the new method is expensive or leads to inconvenience when invasive.
Factors That Influence Sample Size Four parameters influence the sample size: (1) type I error (probability of reje
Data Loading...
