Better health care with better IT, part 2: Implementation challenges

The first posting in this series outlined how Kaiser Permanente has developed a registry of joint replacements as one approach to creating meaningful measures of health care outcomes. This post looks at some of the implementation challenges they, and others seeking to follow their lead, face.

Kaiser is a massive system, delivering care to over 8 million patients, with 350 orthopedic surgeons performing 17,000 total joint replacements each year at 43 medical centers. They've been tracking every joint they've replaced since 2001, meaning they have a lot of data to manage.

1. Electronic Health Record (EHR). The vast amount of data can only be managed and mined using an advanced, integrated EHR. Kaiser uses a system called HealthConnect managed by Epic System Corporation out of Verona Wisconsin. The system unites clinical data with revenue/billing info, healthcare management functions, and applications allowing mobile access and bar code scanning. The data can be queried in a number of ways to extract data for analysis and study, and in the event of a product recall, can quickly report contact information of patients implanted with potentially defective equipment.

2. Developing standardized information gathering, documentation and entry. In the efforts to stream-line data gathering into a efficient, relatively quick process that didn't miss clinically important data, developing questionnaires was a crucial implementation step. Making sure the right stakeholders were involved was also important, both to ensure nothing was missed, and also to build buy-in and support.

3. Unexpected challenges from unexpected areas. Kaiser's effort to use bar code scanning to directly enter the implant information into he medical record, for example, ran into a series of hurdles. Some were technical: Incompatible bar-code language, or inability to scan items not previously entered into the inventory. Some were based on people's habits: The preference of staff to keyboard-enter the data, complexity of maintaining and navigating device look-up tables.

4. Managing the registry required a range of specialized skills, including: Data entry and data quality control; software programmers for data extraction, management, and database maintenance tasks; biostatisticians for data analysis; project managers to provide trainings and chart reviews to validate data. Staff support five specialized orthopedic, and three cardiac registries.

Registries are of course not new. They are best used in Scandinavia, where government run registries in some cases capture virtually every single health care incident that occurs within the country. The Swedish orthopedic registry was the inspiration for Kaiser. Concerns about data privacy and security make it more difficult to implement nation-wide registries in the US. In the US, health care registries can be found in health systems such as Kaiser or Geisinger, and also within certain specialties. Heart surgeons, for example, are well known for keeping track of the details and outcomes of every surgical patient, and their professional society provides guidance and support for these efforts.

The last entry in this series will outline some key learning points from the Kaiser experience.