Statistical consultation can benefit an investigator in a numerous ways, and NBSTRN provides investigators support throughout this process. NBSTRN is staffed with experts in newborn screening and the genetic research fields who are able to assist investigators through the difficulties that may arise when analyzing bio-specimen and genetic data.
Bio-statistical issues may include:
- Choosing statistical methods
- Multivariable modeling
- Other statistical analysis
- Interpreting results
- Responding to reviews
- Sample size planning
- Other bio-statistical issues
Guidelines to Determine Sample Size for Identifying Cases of Rare Disorders
The following tables are presented as guidelines for the sample size from an unselected population to find at least one case (Table 1) or at least 10 cases (Table 2). The population size depends also on the level of confidence required, with greater certainty implying a larger population selection. The results from Table 2 can be scaled for larger numbers of cases. In the case of very rare disorders, the results may suggest that an unselected cohort design is inappropriate based on cost-benefit considerations. In that case, ascertainment of known cases will be crucial for a case-control design.
Table 1: Guidelines for the Sample Size from an Unselected Population to Find at Least One Case
| Confidence (α) | ||||||
|---|---|---|---|---|---|---|
| 50% | 80% | 90% | 95% | 99% | ||
 |
1:100 | 69.3 | 160.9 | 230.3 | 299.6 | 460.5 |
| 1:300 | 207.9 | 482.8 | 690.8 | 898.7 | 1381.6 | |
| 1:1000 | 693.1 | 1609.4 | 2302.6 | 2995.7 | 4605.2 | |
| 1:3000 | 2079.4 | 4828.3 | 6907.8 | 8987.2 | 13815.5 | |
| 1:10,000 | 6931.5 | 16094.4 | 23025.9 | 29957.3 | 46051.7 | |
| 1:30,000 | 20794.4 | 48283.1 | 69077.6 | 89872.0 | 138155.1 | |
| 1:100,000 | 69314.7 | 160943.8 | 230258.5 | 299573.2 | 460517.0 | |
| 1:300,000 | 207944.2 | 482831.4 | 690775.5 | 898716.7 | 1381551.1 | |
| 1:1,000,000 | 693147.2 | 1609437.9 | 2302585.1 | 2995732.3 | 4605170.2 | |
Table 2: Guidelines for the Sample Size from an Unselected Population to Find at Least Ten Cases
| Confidence (α) | ||||||
|---|---|---|---|---|---|---|
| 50% | 80% | 90% | 95% | 99% | ||
|
|
1:100 | 1000.0 | 1302.2 | 1492.6 | 1668.5 | 2047.3 |
| 1:300 | 3000.0 | 3910.0 | 4483.9 | 5014.0 | 6156.2 | |
| 1:1000 | 10000.0 | 13037.3 | 14953.2 | 16723.1 | 20537.3 | |
| 1:3000 | 30000.0 | 39115.5 | 44865.5 | 50177.7 | 61626.1 | |
| 1:10,000 | 100000.0 | 130388.9 | 149558.7 | 167268.7 | 205436.7 | |
| 1:30,000 | 300000.0 | 391170.2 | 448682.2 | 501814.4 | 616324.4 | |
| 1:100,000 | 1000000.0 | 1303904.7 | 1495614.1 | 1672724.6 | 2054431.3 | |
| 1:300,000 | 3000000.0 | 3911717.6 | 4486848.4 | 5018182.2 | 6163308.2 | |
| 1:1,000,000 | 10000000.0 | 13039062.6 | 14956168.2 | 16727283.9 | 20544377.1 | |
If you are interested in statistical consultation or would like more information about resources offered by NBSTRN, please contact us or fill out a NBSTRN Service Request Form.
.png)
Get Adobe Reader