Pitney Bowes Outsourced Predictive Modeling to Protect Privacy

When Pitney Bowes embarked on its predictive-modeling project, corporate privacy policies as well as federal regulations led executives to conclude that the IT department would have to be hands off. Global Health-Care Management Director Jack Mahoney and David Hom, vice president for employment brand total rewards, decided that they had to outsource the entire effort.

Pitney Bowes employees are prohibited from viewing any health information linked to individuals, aside from their own personal information. The rules are designed so that a manager can’t, for example, deny a promotion to an employee because he has a heart condition. Therefore Pitney Bowes can only view its health-care claims data in aggregate.

Because no one in the IT department can have access to individual employee health data, Pitney Bowes has a long-term contract with Medstat, a third-party data warehouse provider, to collect all of its health-care claims and pharmacy data-as well as workers’ comp, clinic, behavioral and disability data.

For the predictive-modeling projects, Medstat normalizes the data and strips it of all identifiers such as names and addresses. The data is then encrypted so that it can’t be traced back to an individual employee. The Pitney Bowes IT department’s role is limited to enabling monthly data feeds from Pitney Bowes’ HR systems to Medstat. As for making aggregate Medstat data available to the desktops of the Pitney Bowes employees who need it, all IT had to do was optimize their browsers.