Flexible Clinical Tolerance Limits for Agreement Assessment
Flexible Clinical Limits for Assessing Agreement by a Direct, Simple, and Robust Method
Dr Abhaya Indrayan,
MSc, MS, PhD (Ohio State), FAMS, FRSS, FSMS, FASc
Former Professor & Head of Biostatistics at Delhi University College of Medical Sciences, Delhi
Biostatistics Consultant at Max Healthcare, New Delhi
Despite several deficiencies (https://sma.org/southern-medical-journal/article/direct-use-of-clinical-tolerance-limits-for-assessing-the-degree-of-agreement-between-two-methods-of-measuring-blood-pressure/ ) the Bland-Altman method continues to be the most common method for assessing agreement between two series of qualitative measurements. The greatest drawbacks of this method are that (i) this requires a Gaussian distribution of the differences and (ii) does not measure the degree of agreement. To overcome these and various other deficiencies in the Bland-Altman method, we have proposed direct assessment by the percentage of differences within the clinical tolerance limits (https://www.preprints.org/manuscript/202108.0343/v1). This method is simple, robust to outliers, non-parametric (distribution-free), and appropriate as it measures the degree of agreement. The confidence interval for the percentage of agreement can be obtained and the tests of hypothesis can be easily done as this is only a proportion.
A significant advantage of our direct method is that the clinical tolerance limits can be flexibly set, such as narrow at critical values and wide at other values (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4189799). For an application of this method for assessing agreement between values of fasting glucose level obtained by two methods, see https://www.sciencedirect.com/science/article/abs/pii/S1871402123001017.
Dr Abhaya Indrayan,
MSc, MS, PhD (Ohio State), FAMS, FRSS, FSMS, FASc
Former Professor & Head of Biostatistics at Delhi University College of Medical Sciences, Delhi
Biostatistics Consultant at Max Healthcare, New Delhi
Despite several deficiencies (https://sma.org/southern-medical-journal/article/direct-use-of-clinical-tolerance-limits-for-assessing-the-degree-of-agreement-between-two-methods-of-measuring-blood-pressure/ ) the Bland-Altman method continues to be the most common method for assessing agreement between two series of qualitative measurements. The greatest drawbacks of this method are that (i) this requires a Gaussian distribution of the differences and (ii) does not measure the degree of agreement. To overcome these and various other deficiencies in the Bland-Altman method, we have proposed direct assessment by the percentage of differences within the clinical tolerance limits (https://www.preprints.org/manuscript/202108.0343/v1). This method is simple, robust to outliers, non-parametric (distribution-free), and appropriate as it measures the degree of agreement. The confidence interval for the percentage of agreement can be obtained and the tests of hypothesis can be easily done as this is only a proportion.
A significant advantage of our direct method is that the clinical tolerance limits can be flexibly set, such as narrow at critical values and wide at other values (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4189799). For an application of this method for assessing agreement between values of fasting glucose level obtained by two methods, see https://www.sciencedirect.com/science/article/abs/pii/S1871402123001017.
