Carnegie Mellon University’s CERT has issued a security advisory on the MEDHOST Perioperative Information Management System (PIMS), a system designed to streamline the progress from patient introduction to surgery. It’s not a vulnerability we should expect these days – especially for a medical system – but PIMS contains hardcoded credentials. These credentials provide access to the customer database.
“An attacker with knowledge of the hard-coded credentials and the ability to communicate directly with the application database server may be able to obtain or modify patient information,” warns the CERT advisory.
“With the surgery department so critical to a hospital’s financial well-being and patient safety initiatives, there’s no room for inefficiencies or costly mistakes,” says the MEDHOST website. …this comprehensive suite of applications fits within existing perioperative workflows, reducing scheduling delays and improving overall patient care.”
MEDHOST has not published any information about this vulnerability. PIMS itself contains several different modules. It isn’t clear from the advisory whether access to the customer database could also provide access to the different modules, such as PIMS Surgery. Nevertheless, the ability to obtain or alter patient data is a serious issue.
MEDHOST has addressed the problem in version PIMS 2015R1 and newer. Administrators are advised to upgrade to the latest version. This is particularly important now that the vulnerability has been made public – malicious actors could reverse earlier versions to obtain the credentials.
The danger with hardcoded credentials is that they can be used to bypass what might otherwise be very strong authentication procedures developed and trusted by the administrators. Since the product can be hosted and accessed remotely, administrators should ensure that only trusted sources can access the server hosting the product. This should be standard procedure whether using a vulnerable version or a newer fixed version.
The credentials do not seem to have been leaked, and it doesn’t appear as if they have been used for malicious purposes. The vulnerability was discovered and reported by Daniel Dunstedter.