My perspective comes from the implementation, clinical content workflow design, UI/UX, database backend, clinical data extraction, and quality reporting side. I've done a lot of EMR training; physicians, nurses, extenders, registration, etc, classroom as well as shadowing clinicians during visits. Here are some quick takes:
Counting clicks -- In my experience, physicians tend to place too much emphasis on counting clicks when evaluating EMR systems. Yes, efficiency (especially in the order and diagnosing workflows) is very important. But avoid the temptation to pick a system based on how many clicks it takes to order an A1c. You'll see a lot of pretty physician screens in the demos. Focus on how easy the system is for your nurses and registration staff to use. That's where inefficiency will kill your ability to keep up with your schedule.
Migration matters most -- If you're converting from a legacy EMR system, one of your primary concerns *must* be data migration. Ask a lot of questions and push for clear answers and costs. What is included? How much historical data will be converted? Has the vendor converted clinical data from your legacy EMR system's schema before? What will *not* be migrated? Will data be migrated at a field-to-field level, or will you only have access to previous encounter documents?
Customization vs out-of-the-box -- Minor customization of EMR settings in order to meet your clinicians' and patients' needs is normal. Every vendor will assure you that you can customize the system to meet your needs. However, extensive customization of EMR screens, flowsheets, data fields, or workflow may prevent successful extraction of patient data for reporting, including HEDIS. Customization may also have a big impact on your ability to take future system updates from the vendor too. (In NextGen's case, any customization of KBM template content will absolutely increase the difficulty of doing KBM updates.) The more you can use out-of-the-box, the better off you'll be in the long run.
Readability of the encounter document -- Pay attention to the format and information contained on the encounter document. I don't just mean whether it's in a traditional SOAP format or whether the HPI reads like a competent physician wrote it and not a chatbot. Consider how the document is going to be used outside of your practice. For example, if a HEDIS auditor is looking for colorectal cancer screening documentation or diabetic eye exam proof, does your document clearly show what procedure was done, by whom, and when? Or does it only show "Colonoscopy -- done"? Auditors and chart reviewers will not have access to your EMR template flowsheets or database fields to confirm that your documentation is accurate and complete. (Trust me, they don't have the time to dig, nor do they have the deep EMR expertise required to research your entries.) They'll be relying on your encounter document. If it's not on your document, you didn't do it. Unfortunately, that's the current environment. "Primary Source Verification" stops at the document, and you'll have to put up with a lot of word salad to pass PSV.
Web-based vs on-site hosting -- Obviously, on-site hosting is going to be more expensive in terms of hardware costs and staff resources than web-based EMR systems. If you can't afford to host your own data, then all of the questions about data migration, ownership, access, and document readability become even more critical. For example, compare Epic hosted to athenahealth hosted. Both vendors can host your data for you and provide a web-based EMR product. But with Epic, you will still be able to work with your data at a field level. Extracting data from athenahealth for quality reporting purposes is much more difficult because the extraction is limited to parsing from the generated text document. If you update a med in Epic, you can report that change directly from the field. If you update a med in athenahealth, you'll have to either regenerate the previous encounter document or wait until the patient's next encounter to make that update accessible for quality reporting.
Epic -- The 800 pound gorilla of EMR systems. This is the most extensible system out there, bar none. If your practice has patient care relationships with a lot of other practices and health care systems on Epic, the advantage of being able to easily share patient data with those organizations and with your patients through MyChart is huge. You will have a lot of help and resources from Epic and other organizations, too. The disadvantages are pretty big for solo practices, though. Epic is expensive, and not just when it comes to initial cost. You will have to dedicate significant staff resources and time for continuing training and certification. (Epic certification is highly desirable, too, so be prepared for the possibility of having your technical staff poached after they get certified.) If cost is not the primary constraint for your practice, then give Epic a serious look.
NextGen -- Next. LOL. This is where I have most of my build experience. NextGen is like a big Lego set. If you have a talented developer on your staff, you can build anything you want. But the backend schema is a mess, the upgrade path for templates and clinical content is extremely hard to manage, and the out-of-box templates require a lot of fixes to work properly. The primary care clinical content is pretty solid, though.
eClinicalWorks -- The business model here is to get clients on low initial cost, and then nickel-and-dime for access to your own EMR data. eCW is popular for a reason, but my concern would be long-term control of your own data. I have had eCW clients complain about hidden or surprise fees, including behavior that borders on outright data blocking. eCW offers a very good utility for extracting clinical data from their eCW Cloud product for reporting. If you have your own talented technical staff, then the utility should be easy for them to use. If you're going to be depending on eCW's staff for everything, then be extra vigilant.
athenahealth -- I would rate the quality of athenahealth's technical support resources pretty highly, maybe second to only Epic's. However, see above for some considerations about reporting data out of athenahealth.
Advanced MD -- An oxymoron.
Other vendors that I've been impressed with from a data extraction POV -- Allscripts, e-MDs
Other vendors to be wary of -- Amazing Charts, Cerner, McKesson. In my experience, all of these have been difficult to deal with from a technical POV.
I hope this is helpful. If you have any questions about what I've posted above, feel free to DM me.
Standard disclaimer: my views are my own and do not represent those of my previous or current employers.
