Validation of individual rules in a
patient-administered digital tool that assesses the need for cancer genetic testing (ASD45)

Callan Russell, Ashley Daley, Durand Van Arnem, Andi Hila, Kiley Johnson, Jill Davies, Colleen Caleshu

Validation of individual rules in a
patient-administered digital tool that assesses the need for cancer genetic testing (ASD45)

Callan Russell, Ashley Daley, Durand Van Arnem, Andi Hila, Kiley Johnson, Jill Davies, Colleen Caleshu

Background

The RISE Risk Assessment Module: Hereditary Cancer is a patient-administered digital tool designed as a reliable, efficient, and scalable approach to identifying individuals in need of evaluation for hereditary cancer predisposition in a variety of settings. The tool’s assessment is based on whether a patient’s personal and/or family history triggers one or more rules in its algorithm, which is informed by national guidelines. Direct validation of the individual rules in the algorithm is needed to ensure the tool’s clinical validity.

Aim

To investigate the clinical validity of individual rules in the algorithm of a digital tool designed to assess the need for genetic testing for hereditary cancer predisposition.

Methods

We compared the tool’s assessment of rules triggered by a case to the same assessment made by the research team. The research team’s assessment was made by a genetic counseling student with supervision by a senior cancer genetic counselor (GC). Cases were selected from our telehealth genetic counseling practice, and were chosen specifically to ensure they would trigger rules that come up frequently in clinical practice. How often rules come up in clinical practice was rated by the supervising senior cancer GC, based on her clinical experience.

Results

Most patients (50/62; 81%) were female and the mean age was 54.3 years old (standard deviation (SD) 13.3). 34/62 (55%) of patients had a personal history of cancer, most often breast cancer (15/34; 44%). The mean number of rules triggered per patient was 4.1 (SD 2.7). There was complete agreement between the research team and the tool for all rules triggered (65/65; 100%). These rules were triggered a total of 269 times, with complete agreement between the tool and the research team each time they were triggered (269/269; 100%). Each rule was validated on a mean of 3.8 cases (SD 3.5). The frequency with which rules are triggered in clinical practice was ranked by a senior cancer GC. All of the rules rated with highest frequency were validated (37/37; 100%), as were most rules rated with intermediate frequency (26/38; 69%).

Conclusion

The complete agreement between the tool and the research team supports that the tool is a clinically valid method of identifying individuals in need of evaluation for hereditary cancer predisposition. The validity of individual rules is particularly critical when applying tools like this in primary care or general population settings since many patients in such settings trigger only a single rule.

Background

The RISE Risk Assessment Module: Hereditary Cancer is a patient-administered digital tool designed as a reliable, efficient, and scalable approach to identifying individuals in need of evaluation for hereditary cancer predisposition in a variety of settings. The tool’s assessment is based on whether a patient’s personal and/or family history triggers one or more rules in its algorithm, which is informed by national guidelines. Direct validation of the individual rules in the algorithm is needed to ensure the tool’s clinical validity.

Aim

To investigate the clinical validity of individual rules in the algorithm of a digital tool designed to assess the need for genetic testing for hereditary cancer predisposition.

Methods

We compared the tool’s assessment of rules triggered by a case to the same assessment made by the research team. The research team’s assessment was made by a genetic counseling student with supervision by a senior cancer genetic counselor (GC). Cases were selected from our telehealth genetic counseling practice, and were chosen specifically to ensure they would trigger rules that come up frequently in clinical practice. How often rules come up in clinical practice was rated by the supervising senior cancer GC, based on her clinical experience.

Results

Most patients (50/62; 81%) were female and the mean age was 54.3 years old (standard deviation (SD) 13.3). 34/62 (55%) of patients had a personal history of cancer, most often breast cancer (15/34; 44%). The mean number of rules triggered per patient was 4.1 (SD 2.7). There was complete agreement between the research team and the tool for all rules triggered (65/65; 100%). These rules were triggered a total of 269 times, with complete agreement between the tool and the research team each time they were triggered (269/269; 100%). Each rule was validated on a mean of 3.8 cases (SD 3.5). The frequency with which rules are triggered in clinical practice was ranked by a senior cancer GC. All of the rules rated with highest frequency were validated (37/37; 100%), as were most rules rated with intermediate frequency (26/38; 69%).

Conclusion

The complete agreement between the tool and the research team supports that the tool is a clinically valid method of identifying individuals in need of evaluation for hereditary cancer predisposition. The validity of individual rules is particularly critical when applying tools like this in primary care or general population settings since many patients in such settings trigger only a single rule.