Some organizations rely on paper-based signatures instead of e-signatures. In both cases, pharmaceutical companies must have the processes in place that confidently ensure that the captured signatures are genuine, are being obtained in the correct manner, and for the correct purpose. Previously, this "confidence" was based on a small random sampling of signatures that were printed and mailed to the signatories with a request that they confirm the printed signature as their own.

The primary flaw in this process is that it is based on a small sampling of the signature database, and that it relies strongly on the signatories' willingness to respond accurately to the confirmation letter. Unfortunately, because physicians now receive a barrage of these confirmation letters from pharmaceutical vendors, the quality of the response rates and resulting accuracy have diminished.

Sam Barclay

For authentication, a signature is an image. With sophisticated pattern recognition technology, the signature image is converted into digital code. The images are then compared mathematically in hundreds of ways, and the outcome is a confidence score. Equally effective in paper and electronic signatures, the confidence score indicates the likelihood that a given signature instance is a forgery or an authentic. To improve the accuracy of the process, several previously captured signatures can be used to create a reference signature, which is in effect an average of a group of signatures.

Any signatures highlighted as suspicious can then be sent for personal review or can be used to create confirmation letters to physicians. Because only suspicious signatures are sent to doctors, fewer letters are generated, and they will be targeted to only those doctors who must respond.

The benefits of electronic signature capture, as well as the ability to store and report on electronic records, are enormous. Some organizations have been reluctant to move off paper-based signatures until the risk of signature fraud can be properly mitigated through technology. Today, advances in e-signature authentication feature accurate character recognition, noise-filtering algorithms, and extraction of key data with extreme accuracy. New technology can also recognize all types of characters, such as machine print, handprint, alphanumeric sequences, international characters, and gray-scale images.

Signature authentication technology improves the complexities of the signature authentication process, increases the confidence in the outcomes and related business and compliance decisions, while enhancing customer experience.

Sam Barclay is vice-president of business development at StayinFront (Fairfield, NJ), tel. 973.461.4800, sbarclay@stayinfront.com