Archive for the tag 'Technical'

Announcing Galen EHR Reporting Webcasts

Galen Healthcare Solutions will be hosting a series of free webcasts covering Allscripts EHR Reporting.  The purpose of these webcasts is to provide insight into reporting options within your EHR database.  We will cover approaches to reporting, database structure, and hands-on querying of the EHR database.

These will be structured in a similar format to university courses – the initial three classes will be at 100, 300 and 500 levels.  The list of the webcasts and their times may be found below.

100 Series – Introduction to the Allscripts EHR Database: Overview of the database, patient demographics and dictionary linking.

  • May 6th, 2009 at 2:00pm EDT

300 Series – v11 Order and Results: querying configuration and patient data.

500 Series – Advanced ConnectR Architecture and Querying

  • July 8th, 2009 at 2:00pm EDT

Please contact Mike Dow, mike.dow@galenhealthcare.com, to sign up.  You must be an existing Allscripts Enterprise EHR client to attend.

We also offer training courses and reporting services for the Allscripts Enterprise EHR database, ETL database, Analytics and the ConnectR  database.  Please contact sales@galenhealthcare.com for more information regarding these courses and our report writing services.

Fun with Problems

I worked with a large group who had been using TouchWorks for a year or so, after converting data from their previous EHR (a predecessor of TouchWorks). The conversion brought their Problems into TouchWorks; however, the problems were viewed differently and were now also viewed and managed by physicians, rather than only the nursing staff. While these changes were generally fine, the personal and family history that they had captured on paper, then showed in their old EHR, now cluttered patients’ problem lists – they had 10 copies of “Coffee Consumption” or “Denies Drug Use”. They also commonly had patients with duplicates of other problems as these didn’t appear as duplicates in the previous EHR.

We spent a few weeks designing a neat de-duplication process. I’ll spare you the details, but it went along the lines of finding problems of the same type, status, category and view. We’d take the oldest entry and keep it. We’d mark the rest as Entered in Error; however, we didn’t toss them away completely – we’d store each problem entry as an assessment on the first problem entry so you could always view the problem details to see the Audit/History.

We also took out some other attributes of the converted problems, such as the category of History Of for every problem and comments that the physician group thought might be useful.

Before I continue, I have to say – both the physician group and those assisting with the conversion (at Allscripts) did a great job of converting the 12 years of electronic charts (about 2MM patients). I think the biggest issue was the group not having a year’s experience using TouchWorks to make some of the decisions – hindsight is 20/20.

Luckily, we had the ability to make corrections as we went. We were able to convert millions of problems, removing duplicates (while maintaining their history!) and saving the clinicians a great amount of time and frustration. An internal medicine doc remarked that it saved him somewhere in the area of 100 hours of effort (not to mention the gray hairs).

I wouldn’t say we did anything extraordinary, but just had the luck to work on a fun project that helped out a couple hundred nurses and physicians. And hopefully make their transition to a (full) EHR a little more pleasant.

A Better Training System

Groups using the Allscripts Enterprise EHR (formerly TouchWorks EHR) that I’ve worked with have always yearned for a better training system. Issues with the current training system run the gambit, starting with the majority of groups using their Test or Production/Live system for training purposes. The groups that do have a training system often have a number of very reasonable requests.
Let me outline the ones I’ve always seen:

- The training system build / configuration isn’t close to production,
- Or it isn’t close to the test/development environment where new versions or new build work is available.
- The patient data isn’t realistic.
- The “test” patients become “cluttered”, e.g. 100 problems, 500 notes, 300 medications, etc, etc.
- The patients in the training system are real patients.
- The information is out of date. For example: tasks are months or years overdue or there are no appointments on providers’ schedules.
- If any of these are accomplished (for example, Galen has worked with a handful of groups to add some of the smaller, items such as de-identifying real patients, updating schedules and task lists), the physician group finds that changes are difficult, such as training both new users on the current version, and preparing to train users on a new version that they are upgrading to soon.

There are some serious technical hurdles to overcome to have a training system that avoids the above pitfalls and any others I failed to mention. The biggest hurdles come in combining an EHR “build” (menus, Note definitions, tasking work flows, etc) with patient data. If you can do that successfully, there may be other hurdles – for example, I talked to a large group about a training system several months back. We were able to design it in such a way that we overcame all of the above issues, but it required the client to purchase another server and work through some serious red-tape with their IT group, especially as it related to copying their production system. They would have been interested in the solution, if it were reasonably feasible to get a system that could do everything outlined above, but was completely remote – e.g. I would pick up their build (say from their prod or test system), upload it to a TouchWorks environment that we hosted, then play in the clinical / content information, such as patients, their charts, tasks, appointments, etc.

I’m curious to hear what folks think about training environments, what’s missing and what could be done without. Ideas?

Panel Query

Panel Query in TouchWorks promised to make all the clinical data accessible to the everyday provider. Anyone could just pop into Panel Query, write a “query” based on an easy-to-use interface. No need to know SQL or any programming. And to protect the system, the queries were to be scheduled, presumably overnight.

TouchWorks largely delivered on this promise. Panel Query certainly delivered on the first part – anyone who is already using TouchWorks, can easily build usable reports with no technical/SQL background. Panel Query does not make every piece of information available, but it’s a great start. The problem lies in the execution of the query. First there is the performance aspect (Panel Query can really be a drag on your TouchWorks system’s performance), you also need to schedule the report, then the results come back as a patient list, which can be difficult to work from.

Performance is enough of an issue for most TouchWorks clients, adding queries that hit the system hard definitely won’t help things! Most practices can allow queries to run during the night, but Panel Query doesn’t enforce this – it allows you to schedule queries whenever you’d like. Other practices, say those with 24/7 urgent care clinics, cannot afford to have queries running during the night.

I worked with a group recently who had a copy of their production TouchWorks database (refreshed regularly) on their data warehouse server. They came up with an idea — could TouchWorks be setup to overcome the issues they had with Panel Query. Could Panel Query:

  1. Point to their data warehouse
  2. Make the results available in an Excel spreadsheet
  3. Execute immediately – take out the additional steps of creating a job and scheduling it.

We’ve done all three, now anyone with access to Panel Query, can run queries while they are logged in to their production TouchWorks system (Panel Query then re-points to the data warehouse) any time of day and get their results back immediately.

I bring it up as it was a fun project that I really enjoyed working on. That’s all. What other projects have you folks worked on? Any ideas for other projects like this that you haven’t done?

« Previous Page