Breach Notification for Unsecured Protected Health Information
Posted May 06 2010 12:00am
By: Michael R. Spaltro
Gordon Moore, Intel co-founder, famously predicted that the speed of technology will double about every two years. Between 1981 and 1991 , “computer processing speed increased tenfold, the instruction execution rate a hundred fold, system memory grew a thousand times, and system storage expanded by a factor of 10,000.” That was just the beginning. Intel has kept that pace for nearly 40 years, now introducing the world’s first 2-billion transistor microprocessor. The development of fundamental computer technology has translated into ubiquitous information technology infrastructure. Deploying information technology within the healthcare industry is significantly complicated by the indispensability of life and health to everything else we do. The privacy of electronic health records (”EHR”) that contain personally identifiable health information (”PHI”) is one area of particular concern.
Health care providers, health care plans, health care clearinghouses, and their business associates across the country are currently using EHRs as an efficient method to locally store patient records.  EHRs may contain patient treatment history, social and demographic data, and a multitude of other personal health information (”PHI”).  If the underlying computer technology continues to grow at the staggering pace predicted by Moore’s Law, the function of EHRs will expand to “assume a key roll in medical diagnosis and treatment management.”  Moreover, the Food and Drug Administration, in collaboration with public, academic, and private entities, is expected to use EHRs to link and analyze medical safety data from over 100 million patients by July 2012.  The resulting electronic network of interoperable healthcare data is of a scale never before contemplated in the industry. Personally identifiable health information, such as the data contained across local provider EHRs, health plan claims databases, and Medicare databases, will be remotely transmitted, stored, accessed, and analyzed.
Transmitting EHRs between an originating entity and the entity/infrastructure involved in research, development, and storage of EHRs, creates an increased potential for internal and external breach. Moreover, as EHRs become populated in local and remote institutions across the country, the incidence of breach ostensibly increases. In the event of breach, an individual may be exposed to a number of dangers. EHRs contain personal information of high value to computer hackers, such as social security numbers or payment information.  Furthermore, an otherwise legitimate entity could potentially use health information in a less nefarious way that nonetheless breaches individual privacy. How can we legally protect privacy while realizing the benefit of electronic health information technology?
The Health Insurance Portability and Accountability Act (”HIPAA”) shores up unauthorized access to protected health information. The HIPAA Security Rule and Privacy Rule require an entity such as a health plan, health care provider, business associate, or a health care clearinghouse, to safeguard all protected health information. Civil and criminal penalties are enforced against entities that fail to comply. The FDA’s qualified contractors  will similarly be subject to HIPAA under the Health Information Technology for Economic and Clinical Health (”HITECH”) Act by 2017.  Therefore, the entire electronic network of EHRs will be covered by the Privacy Rule and the Security Rule. Within covered entities, protected health information is to be stored with any security measure that allows an entity to reasonably and appropriately implement all safeguard requirements. The Security Rule approves that a covered entity may use firewalls and other access controls (such as passwords) to safeguard PHI in its electronic form. Without this intangible structure protecting EHRs, unauthorized parties could easily access PHI and PHI could easily flow out to any individual, device, or system that interoperates with EHR databases. The HIPAA Security Rule therefore assures that a covered entity is reasonably protecting an individual’s privacy by safeguarding personal health information.
Firewalls and other reasonable access controls are not impermeable. Earlier this year, an ultra sophisticated hack attack on Google penetrated the multi-billion dollar corporation, causing it to later withdraw from China. Merck & Co. and Cardinal Health Inc. were among others infiltrated in the attack. The extent of information exposed is still not fully understood. Thus, breaches occur even if reasonable and appropriate safeguards are required. The access controls required by HIPAA in the Security Rule are not sufficient to protect a vast network of interoperable EHRs. Further data encryption and/or secure data destruction will eventually be required to protect individual privacy.
Pursuant to the Privacy section of the HITECH Act, Title XIII Division A, Subtitle D, the Department of Health and Human Services (”HHS”) was required to promulgate breach notification for unsecured protected health information rules and regulations (”Breach Rule”). HHS issued a final rule, effective September 23, 2009, requiring all entities and business associates covered under HIPAA to provide notification in the cases of breaches of unsecured protected health information. Presumably, an individual who is made aware that his personal information was compromised is better equipped to mitigate identity theft or other harms that could arise.
The provisions in Section 13402 of the HITECH Act are consistent with HIPAA definitions of a “covered entity” and “protected health information.” The Act defines breach as the unauthorized acquisition, access, use, or disclosure of protected health information which compromises the security of that information. In other words, if a firewall or reasonably appropriate access control is breached a covered entity must report that breach to all of the individuals affected. Importantly, notification of breach is only required for unsecured personal health information. If a covered entity is in the practice of encrypting and/or destroying PHI in accordance with the National Institute of Standards and Technology (NIST), then that entity does not have to report a breach of their firewalls or access controls. It is only necessary to provide notice if “unsecured protected health information that is not secured through the use of technology or methodology specified…” is breached. The rationale is obvious. If a covered entity encrypts PHI in accordance with NIST standards, then the data is unusable in the event of a breach, and notification would be superfluous.
Consequently, a covered entity has two choices: (1) secure all EHRs that contain PHI; or (2) report breaches of PHI. The Breach Rule encourages cover entities to take the former approach. To secure EHRs that contain PHI, an entity must regularly perform two standard procedures. First, the NIST published standards recommend a “one pass” method of data deletion for most applications.  When electronic data is deleted, it is only removed from the file system. The “image” of the data physically remains on the hard drive of the device. Software and hardware methods of recovering deleted data are available to the public. Therefore, “deleted” PHI data could be recovered by an unauthorized entity in the event of a breach. The NIST recommends that one data overwrite be performed on the deleted data, as to render it unrecoverable. Depending on the method used and size of the database, data deletion can take up to an hour.
Second, and perhaps less straight forward, the NIST recommends data encryption using one the following four methods: full disk encryption; volume encryption; virtual disk encryption; or file/folder encryption.  The capital expenditure necessary to install and maintain encryption software/hardware throughout a covered entity is immense. Furthermore, encrypting millions of EMRs will tax computer processors and networks, and will additionally hamper interoperability. When data is encrypted it losses all functionality, and therefore must be decrypted by the authorized end-user before each use. It would be additionally problematic to transfer encrypted data throughout an electronic network, like that contemplated by the FDA, unless all systems were equip to recognize and decrypt the data. Thus, under either of the encryption methods above, the net result is a loss of productivity and interoperability. Moreover, encrypted data may not be mean secure data. The end-user authorized to access encrypted data will likely decrypt it during the course of a work day. Therefore, so-called encrypted PHI would be exposed to the same daily risks as unsecured PHI. Consequently, the nature of data encryption may not even provide the security and privacy that the Breach Rule contemplates.
While some covered entities are voluntarily choosing to encrypt and secure PHI, the impracticality and cost of data encryption is prohibitive. Covered entities were allowed 180 days to become compliant with the Breach Rule. That period has expired, and most covered entities have not opted to encrypt PHI. Instead, covered entities have put reasonable systems in place to detect breaches, as required by the Breach Rule. The Breach Rule requires notification without unreasonable delay once a covered entity learns of a breach. A majority of states already had breach notification laws in place, and thus covered entities had respective systems in place to detect and report breaches.
Reporting breaches under the Breach Rule still requires some capital expenditure. In some cases, notification to popular media outlets and the Secretary is required. This notification could potentially detract business and invite legal action. Of greater concern, a major breach and broadcast resulting in legal action may dissuade industry players from adopting EHR systems that could potentially reduce medical error and healthcare costs.  However, the burden of encrypting PHI is overwhelming, and perhaps ultimately ineffective. Consequently, the Breach Rule has done little to foster the actual security of PHI. In practice, covered entities merely provide notification of breach. It is unclear how this may or may not benefit a patient whose privacy has been breached. Deploying new EHR technology throughout the healthcare industry presents a risk to individual privacy that is not adequately addressed by the Breach Rule and HIPAA.
Privacy concerns should positively correlate with the volume of online EMRs. Pursuant to the FDAAA, 100 million EHRs will be linked within the FDA’s seminal network by July 2012. The sensitive and valuable nature of robust EHR databases will likely attract the attention of unauthorized parties around the world, and should therefore warrant a heightened level of security. Within two years, encryption technology may prove to be significantly smarter, cheaper, and more efficient. The concerns that bar covered entities from adopting data encryption may be lifted. While absolute data security is not likely attainable under any standard, software operating systems that integrate on-the-fly encryption would be ideal and foolproof. Rules and regulations should proportionately reflect advances in computer technology and the quantity of EMRs over the next two years. To protect public privacy and trust in our healthcare system, all PHI should eventually be encrypted by covered entities and their business associates.
 Hoffman and Podgurski, Finding a Cure: The Case for Regulation and Oversight of Electronic Health Record Systems, 22 Harv. J. L. & Tech 103.