A scoping review of statistical methods in studies of biomarker-related treatment heterogeneity for breast cancer

Table 2 Characteristics of statistical analysis of selected studies of biomarker-based treatment heterogeneity

Characteristic	Number (%^a)
Type of analysis
Cox model with biomarker-treatment interaction^b	16 (52)
Logistic model with biomarker-treatment interaction^b	6 (19)
Subgroup analysis^b
Treatment effect by biomarker subgroup	20 (65)
Biomarker effect by treatment subgroup	7 (23)
All biomarker-treatment subgroups	1 (3)
Subpopulation Treatment Effect Pattern Plot	2 (6)
Bayesian covariate-adjusted logistic model	1 (3)
Difference in cumulative incidence at one time point	1 (3)
Test of interaction
Yes	22 (71)
No	9 (29)
At least one significant result
Yes	21 (68)
No	10 (32)

^a Percentages do not sum up to 100, if a study qualified for several categories or due to rounding
^b One study presented interaction models for binary as well as time-to-event endpoints, and additionally biomarker effects by treatment subgroup and treatment effects by biomarker subgroup and is included in more than one category

ISSN: 1471-2288