December 12, 2013
There is less than a year until ICD-10 implementation, and if you aren’t currently deep within your own implementation process, you may be starting to...
There is less than a year until ICD-10 implementation, and if you aren’t currently deep within your own implementation process, you may be starting to feel like the pressure is mounting. And honestly, you aren’t entirely wrong. There are certainly a good amount of steps that should already have been completed, but don’t worry; ten months is a long time to clean up loose ends.
Or maybe you’re having the opposite reaction: ten months should be plenty of time to complete testing, right? How complicated can it be? As testers, we run into this mindset a lot, and unfortunately it’s incredibly misguided. The testing effort should be considered almost as important as the development process, especially with a system which has the potential to make things really difficult for a huge number of people.
Let’s back up a second, though, because anyone who isn’t in the healthcare IT industry is probably really confused right now. ICD-10 refers to the new medical coding system created by the World Health Organization to replace the now-outdated ICD-9 coding system. The US government has mandated that healthcare organizations must migrate their systems to the new coding protocol by October of 2015, which gives those organizations about 10 months to implement it. Failure to do so compromises their ability to serve their patients and work with insurance companies; however, even if you aren’t in healthcare, preparing for ICD-10 implementation serves as an excellent case study to gain insight into the testing lifecycle for complex migration projects.
Testing for ICD-10 in particular is not only hugely influential for everyone who works with a healthcare organization in any capacity; it’s also incredibly detailed and complex. According to the American Medical Association, organizations should budget 2-3 months for internal testing and an additional 6-9 months for external testing. That puts the testing life cycle at 8 months at best and one year at most; this means that if testing has not yet begun, meeting that October 2015 deadline is becoming increasingly unlikely.
Worried about your ICD-10 testing effort? We’ll try to calm your fears by outlining what you should already have done, and what other undertakings you ought to begin soon.
Where should your ICD-10 implementation be with 10 months left?
What most organizations have already completed:
- Test assessment and planning
- Budget creation and approval
What probably still needs to be tackled:
- Revenue impact testing
- External integration testing
- Determine how they will get their test data: coding it themselves, using cross-walks/GEMS, or acquiring a tailored dual-coded database
- Determine procedure for clinical documentation improvement
- Decide whether to automate any portions of their testing effort, and determine which tools/products will be used in the process
For starters, it’s like that you’ve already finished your test planning and assessment phase – and if so, good on you! This necessary first step provides a roadmap for what you will need to do and when over the next ten months; not doing so will leave organizations without information that’s vital for the testing process. Luckily, our survey shows that around 57% of organizations have already completed their test assessment, and another 31% are currently in the process, so only the remaining 12% should be worried about this one. It would also be beneficial to have a budget in place and approved by this point.
Part of what the assessment helps with is determining the nitty-gritty details of how your organization will generate its test data. The typical options either use cross-walks/GEMs or an independent dual-coded database. Typically the latter is recommended, as it’s been found to be more reliable. And then, even once you’ve determined which works better for your organization, there’s more to ask: Will you be setting this system up yourself, or using one that was compiled and tested by another company?
The next important step to begin is revenue impact testing and external integration with other payers/providers. In addition to ensuring that your system works internally, it is vital to ensure that it communicates efficiently with systems in other organizations and businesses as well – at least, it’s vital if you’re planning on getting paid (which we think is a safe assumption). This is also a good point to determine what your procedure for clinical document improvements, or CDI, will be, but this step is not quite necessary at this point.
Ten months in advance is also a good time to start determining whether or not you will automate parts of the testing effort, or if you will manually conduct the entire process. It is an important point to consider at this point, because, again, this decision opens up a whole ‘nother can of questions that need answering, like: which processes will be most fruitful to automate? How long will that process take? What products will be required, and can they be used afterwards as well? We recommend a good amount of automation with such a massive testing effort, because it reduces the risk of human error; and when you consider how much is riding on proper ICD-10 implementation, with a possible revenue impact on the organization as a whole, you really want that risk to be as small as possible.
Does ten months still seem like a really long time to complete testing? Now hopefully you can see why it’s incredibly important that you initiate the testing process for your ICD-10 migration ASAP, if you haven’t yet, and why it’s so important for your organization to keep on schedule if you have!