Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance
Objectives: To quantify variability in the results of statistical analyses based on contingency tables and discuss the implications for the choice of sample size for studies that derive clinical prediction rules. Study Design and Setting: An analysis of three pre-existing sets of large cohort data (...
| Main Authors: | , , , , |
|---|---|
| Format: | Journal Article |
| Published: |
Elsevier
2017
|
| Online Access: | http://hdl.handle.net/20.500.11937/54955 |
| _version_ | 1848759506713968640 |
|---|---|
| author | Kent, Peter Boyle, E. Keating, J. Albert, H. Hartvigsen, J. |
| author_facet | Kent, Peter Boyle, E. Keating, J. Albert, H. Hartvigsen, J. |
| author_sort | Kent, Peter |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | Objectives: To quantify variability in the results of statistical analyses based on contingency tables and discuss the implications for the choice of sample size for studies that derive clinical prediction rules. Study Design and Setting: An analysis of three pre-existing sets of large cohort data (n = 4,062–8,674) was performed. In each data set, repeated random sampling of various sample sizes, from n = 100 up to n = 2,000, was performed 100 times at each sample size and the variability in estimates of sensitivity, specificity, positive and negative likelihood ratios, posttest probabilities, odds ratios, and risk/prevalence ratios for each sample size was calculated. Results: There were very wide, and statistically significant, differences in estimates derived from contingency tables from the same data set when calculated in sample sizes below 400 people, and typically, this variability stabilized in samples of 400–600 people. Although estimates of prevalence also varied significantly in samples below 600 people, that relationship only explains a small component of the variability in these statistical parameters. Conclusion: To reduce sample-specific variability, contingency tables should consist of 400 participants or more when used to derive clinical prediction rules or test their performance. |
| first_indexed | 2025-11-14T10:00:58Z |
| format | Journal Article |
| id | curtin-20.500.11937-54955 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T10:00:58Z |
| publishDate | 2017 |
| publisher | Elsevier |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-549552018-06-06T07:01:17Z Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance Kent, Peter Boyle, E. Keating, J. Albert, H. Hartvigsen, J. Objectives: To quantify variability in the results of statistical analyses based on contingency tables and discuss the implications for the choice of sample size for studies that derive clinical prediction rules. Study Design and Setting: An analysis of three pre-existing sets of large cohort data (n = 4,062–8,674) was performed. In each data set, repeated random sampling of various sample sizes, from n = 100 up to n = 2,000, was performed 100 times at each sample size and the variability in estimates of sensitivity, specificity, positive and negative likelihood ratios, posttest probabilities, odds ratios, and risk/prevalence ratios for each sample size was calculated. Results: There were very wide, and statistically significant, differences in estimates derived from contingency tables from the same data set when calculated in sample sizes below 400 people, and typically, this variability stabilized in samples of 400–600 people. Although estimates of prevalence also varied significantly in samples below 600 people, that relationship only explains a small component of the variability in these statistical parameters. Conclusion: To reduce sample-specific variability, contingency tables should consist of 400 participants or more when used to derive clinical prediction rules or test their performance. 2017 Journal Article http://hdl.handle.net/20.500.11937/54955 10.1016/j.jclinepi.2016.10.004 Elsevier fulltext |
| spellingShingle | Kent, Peter Boyle, E. Keating, J. Albert, H. Hartvigsen, J. Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title | Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title_full | Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title_fullStr | Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title_full_unstemmed | Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title_short | Four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| title_sort | four hundred or more participants needed for stable contingency table estimates of clinical prediction rule performance |
| url | http://hdl.handle.net/20.500.11937/54955 |