Estimating the potential for data misuse is essential for all data-sharing decisions. This work presents $${tkl}\hbox {-}\mathrm {Score}$$tkl-Score which extends the state of the art $${M}\hbox {-}\mathrm {Score}$$M-Score and $${L}\hbox {-}\mathrm {Severity}$$L-Severity measures. The new proposed measure is sensitive to the increased misuse potential when records are more identifiable in a source table with $${l}\hbox {-}\mathrm {Distinguishing~Factor}$$l-DistinguishingFactor and also when sensitive attributes are less granular in a source table with $${l}\hbox {-}\mathrm {Distinguishing~Factor}$$l-DistinguishingFactor and $${t}\hbox {-}\mathrm {Distinguishing~Factor}$$t-DistinguishingFactor; in contrast, the earlier $${M}\hbox {-}\mathrm {Score}$$M-Score and $${L}\hbox {-}\mathrm {Severity}$$L-Severity only account for record identifiability in a source table with $${k}\hbox {-}\mathrm {Distinguishing~Factor}$$k-DistinguishingFactor. $${tkl}\hbox {-}\mathrm {Score}$$tkl-Score is shown to better characterize the risk of releasing records compared to $${M}\hbox {-}\mathrm {Score}$$M-Score and $${L}\hbox {-}\mathrm {Severity}$$L-Severity due to accounting for sensitive attribute granularity.