In Search of Document Management Software

This blog is a technical discussion for people that need to make IT decisions about document management. If you don't care about IT, have never heard of a database or don't need to mange documents I suggest you go have some fun at this web page or this one.

If on the other hand, you've ever grappled with the problem of what to do with that electronic report or fax this blog's for you. Here's the problem. Not all EMR software is intended to manage masses of documents. Most are based on small databases not intended for dozens of users or masses of images (e.g. Access). Larger databases can handle scanned images and many users such as SQL Server, SQL Anywhere or Oracle. One of the main differences is how they handle errors such as when a connection is lost when updating the data. In smaller databases, an error like this can corrupt the entire database. This will become important in a minute.

The smaller document management solutions allow you to scan the document, the program then creates a file folder with an automatic name and saves the new file into that file folder. To relate the unique patient ID to the file folder the programs usually maintain a small database. If this database gets corrupted all links to the file folders are lost and there is no way to figure out which patient belongs to which file folder.

The better way to do it is to store images into a large database. Our offices' EMR is based on SQL Server. It can handle large scanned images (like x-rays and scanned documents) but the 'front end' is inadequate, matching it to TWAIN compliant machines (scanners, camera) is expensive and adding scanned images or documents to it would bloat the database.

The answer is to find a SQL Server based document management software with a TWAIN driver interface that is not intended for a massive company. Opentext and AGFA solutions are more than robust enough but too expensive for our office.

Luckily, smaller companies are starting use enterprise databases (like SQL Server) which are exponentially safer for critical data. The latest company that we're looking at is Dolphin Imaging. It started as a dental product for managing dental x-rays, pictures and scanned documents but seems to have evolved into a low cost alternative for document management. The major benefit is it handles most TWAIN devices and saves the image files to SQL Server.

I've found out the hard way that how document management software stores the images and data is far more important that what the interface looks like. I would warn everyone that before you start scanning all of your patients faxes, lab reports and requisitions into an EMR you check under the hood.

EMR Flame-Out

Michelle Griever is a family physician in Toronto, Ontario who supports a large family health team. They received funding 2 1/2 years ago from the government to implement an EMR. She's blogged about her journey.

After 2 1/2 years only 2/3rd of the physicians have adopted the system. Maintenance does not seem to be sustainable and they are still deciding on whether to keep control of the servers in-house or have them housed off-site. In short, her practice is hitting a road block that completely derail using EMR for anything more than billing, booking and lab work.

The story underscores an important point. The cost of maintenance, replacement and network design. When our office created the network, we gave little thought to the cost of network architecture and maintenance. The reality is that it (and bandwidth) can make a sizable portion of the total IT budget. The typical cost of an 'entire' system (back bone, servers, bandwidth and workstations) can easily reach $10,000-15,000 per work station for set-up. Clinicians often focus on the utility and cost of EMR without seeing the 'hidden' costs.

We are now 5 years into implementation and still not paperless. But, our system is stable, sustainable and has a calculable return on investment (of about 1 year).

Bring an EHR to Every Canadian by 2016

Infoway Canada (a collaborative of provincial Health Ministers and Federal funding) announced greater funding for electronic health records and the goal of an electronic health record for every Canadian by 2016.

I'm conflicted by the report. On the one hand, electronic records can help patient care. The infrastucture is needed (just the other night a radiologist 400km away read an emergency CT for me) and I'm a believer in EHR. I don't believe they will revolutionize health care though. On the other hand, it looks like we have a greater chance of having and electronic health record for every Canadian before we have a doctor for every Canadian. Are we putting the cart before the horse?

Network or Application - Which is more important?

Check out any EMR resource and you'll find tons of data on how quickly the EMR program responds. Unless you're a super-specialist, however, you probably see more than 10 people a day and the amount of time you spend in the EMR software itself is a few minutes per patient. Some of the major causes of headaches come not from the EMR software itself but network issues.

That means that the impact of printer issues, log-on, log-off and other network associated delays can chew up a serious amount of time (and anger) each time a user accesses the EMR software. For a fantastic description of a day in hell due exclusively to network issues check out FatDoctors post.

I am not trying to advocate against EMR (in fact I'm strongly for it) but when planning a deployment network design should feature just as prominently in the decision analysis as the type of EMR software. In our office, we had the network up and running for just over a year before we deployed the EMR. The goal for time on the computer should be about 10% of the total appointment time including any network issues (log-ins, printer choices, etc...) that need to be adjusted.

More blogs on InPatient Documentation

John at Life as a Healthcare CIO also has a post today about documentation. See what one of the most progressive hospitals in the US is planning. He's posted the minutes from their retreat on inpatient documentation.

EMR Patient Notes - Lessons Learned

The single biggest push back I've had to a completely digital office is from the providers who have to make patient notes. Many cannot type well and all are bankrupt of time. Here are the lessons I've learned from trying to convert people:

1. Finding a patient needs to be easy. Making notes is done from the calender (right click on the patient and hit Add>>Note). Making the providers hunt for a patient by name or number is too time consuming. There needs to be a direct link from the schedule.
2. Total time counts. The providers don't care about login times, response times, printing times. They only care about the total time to takes to get into the system, make a note and get out. Further, it is proportional to the length of the appointment. I find they will tolerate making a note for 10% of the total appointment time. If it is a very short appointment they want an instant note (drop down notes) and if it is an exceptionally long note they will want to dictate it.
3. Logging in takes time. Originally we would have the provider log into the computer than the EMR software so that the active directory could be used for authentication. It was too time consuming so we switched to a 'kiosk' system where the provider just has to log into the EMR software
4. Some things have to be dictated. There are some notes that need to be dictated. Either the provider is a slow typist, the note is detailed, long or requires a lot of thought. It should be as easy to dictate as type a note.
5. Templates are great if they are personalized. We use Word templates with data merged from the EMR system and Fill-In boxes. Most of the providers prefer to learn how to modify the template and make their own.

Of all of the road-blocks I've run into time is the biggest. Keeping the start-to-finish time to make a note about 10% of the total appointment time is critical. That means having different ways to make a note. Boiler plate text, templates, free-hand and dictation all play a role to achieve the 10% goal.

Haven’t adopted EMR yet? You’re not alone.

As recently reported in the New England Journal of Medicine only 4% of doctors in the US have a fully functional electronic records system and most of those are in massive group practices. With the supposed benefits and battle cries for patient centric care one would assume that the demand would be greater and the adaptation quicker.

To my eye the demand is not there yet;

• In my own practice fewer than 5% of referrals are electronic, 1/20 patients provide an email address and 1/50 list it as their preferred method of contact (all are offered). It is hardly an overwhelming demand for electronic communication.
• When done well, efficiency does not change and if implemented poorly it can lower productivity.
• The cost of installation can easily reach $10,000 per seat and the maintenance fees $1,000 per year per seat.
  

So why switch?

• There is less chance for error especially with follow-up such as with labs, other tests, referrals and preventive care. With EMR the chance of forgetting certain items drops to nearly zero.
• Increased capacity. Even though productivity can drop in a poorly implemented system, (e.g. daily billings) capacity increases by eliminating repeated or unnecessary work. In our office we saw an 18% improvement in capacity the first year and 6% the next.
  

Most physicians that participated in the survey also stated they planned to make the switch soon. Improved networks, a greater emphasis on efficiency by programmers, government incentives and diminishing cost for wide bandwidth will all make the transition easier and not too far in the future.

Electronic Record Dictionary

Confused at the difference between an EMR, EHR, PHR and all other electronic acronyms? Check out Life as a Healthcare CIO for a post to de-jargonize the alphabet soup of terms.

The New RIM Blackberry Bold and Primary Care

Living in the home town of Research in Motion (maker of the Blackberry) you get a little treat every now and again. One of them came in 1999 when someone told me to buy stock in the company (which I of course ignored) and the other was to see the new Blackberry Bold last week. The only way to describe the Bold is a cross between an iPhone and traditional Blackberry. It’s got a high resolution screen, keyboard, video, camera, music player, Wi-Fi, GPS and every feature you can imagine. Since 1999 I’ve owned five different Blackberrys and based on my initial impression this one is going to change commerce. The Bold looks like it could be the candidate that will merge the business and consumer applications. Before people reading this blog think I’ve been paid to do a promotion, I’ll get to the point of the article. I and many others have posted about the inefficiencies of EMR and digital note keeping. How clinicians often stop writing notes because it takes them longer to get into and out of the notes section than to see the patient. The Bold’s design, high resolution screen and connectivity has the potential to be able to view a days schedule on Blackberry, pull up the patient the way you normally would and make notes. Because it would be your personal Blackberry, log-in and log-out could be authenticated through the BES server (which connects your corporate email to the Blackberry network) and available on your hip. It has the potential to replace a computer in every room.

Nothing is so easy, however, because pushing data out is one thing (which RIM has mastered) but sending it back is quite another. Part of the problem, as I understand, has to do with authentication. In a digital nervous system there exists an active directory of user names and passwords which grant permissions' to users over a network. This is the reason you can log onto any computer in a clinic without loading your user information first. Blackberry’s would need to authenticate the specific patient entry, send it back through the BES server, into the network and finally into the clinical software. To complicate matters the clinical software often requires its own password. To create a secure zone between your Blackberry and the database that manages patient data is not simple; for now.

With the launch of the Blackberry Bold, RIM also announced the release of their BES Server 5.0 which has something called Blackberry Administrator Service (BAS) and Unified Application Management. Simplified, it means that the BES server may be able to turn your Blackberry into the equivalent of a desktop computer with all of the security and functionality of a wired network but available over a WAN.

The reason I bring this up for smaller clinics is that the cost of EMR is not small and the ROI dubious. RIM has been involved in hospitals but I have not heard of them pushing data into and out of small databases like they do with Exchange. As I found out the hard way, the greatest cost of ownership is not the EMR package but the network and peripherals. One of the reasons our own clinic hasn’t changed to tablets is that both the hardware (tablets & Wi-Fi) and maintenance of a wireless database network is expensive.


Providing a Blackberry to clinicians over a Wi-Fi or WAN is much less expensive than installing and maintaining your own complete network. Plus, RIM has already mastered secure data transmission. If possible, a BES server would simply be installed next to the EMR server and clinicians could access and enter information immediately. Like with any IT roll-out the project should improve quality and efficiency. If it only does one or the other it will be an up-hill battle. If it does neither, the project is doomed. RIM BES Server 5.0 and the Blackberry Bold have the potential to capitalize on the database changes that are occurring because of Health 2.0. Assuming that the back end of Blackberry can be synchronized the Bold looks to be a promising front end which would provide the quality of the tablet in a more cost-effective and efficient manner.

Available Records

Today it is time to boost the virtues of electronic medical records (EMR). I may not believe that posting the symptoms of your abdominal pain on Facebook will help you get a better diagnosis nor do I think that EMR is the saving grace for out-of-control health care wait times but I will say this; EMR increases the ease of access to your medical record. If you don't think it is a problem check out this article by the BBC on the state of health records in NHS clinics (UK).

“The HSJ [Health Service Journal] found six out of the 49 trusts had 5% or more notes missing. On average 2.6% of outpatient records were missing.
City Hospitals Sunderland NHS Foundation Trust reported the highest rate, with 19% of outpatient records unavailable at the start of clinic.”

Those are some crazy numbers. Let’s assume that the average general practitioner sees 30 patients over the course of a day. By extending the numbers that means at the start of the clinic 6 of the patients won’t have a chart present, 2 notes will be forgotten (missing) and 1 patient visit won’t be recorded at all.

I will say this for EMR and electronic scheduling; they improve practice management. A record of every appointment is automatically created, no charts will be missing (short of recording the wrong patient name/identifier; but that’s another post) and it is easy to make some sort of a rudimentary note. It’s been my experience that when a clinic is hopping, providers will not make notes whether they are using a paper or digital chart. So missing notes will always be a problem. But, you can create logic in the EMR (or have a staff member double check the computer) to ensure every patient that arrived had a note made by the clinician who cared for them. EMR may not save health care wait times but it will definitely decrease on paperwork chasing.

For a blog-en-blog throw-down about e-Health in general go to Dr. Wes it’s a great read.

The Antithesis of EMR

Adventures in Medicine posted yet another great blog last week on where charting has gone. As you sort throught the EMR landscape and review the details of blood pressures from 3 years ago on you're patients consider this. They found a chart dating back 50 years with a simple entry:

"Hospitalized for bacterial meningitis. Fully recovered"


Do you really need to know anything more?