When doing an upgrade of any software, there should be some basic steps taken to protect yourself and your office. I don't know if there are documented on Jon's website but some basics of change control include:
1. Do the upgrade on test hardware or in a virtual machine environment. This makes sure you are comfortable with all aspects of the upgrade and have documented what you need to be successful.
2. Give yourself only a certain amount of time to do the upgrade, the change window (say Friday 6:00pm thru Sunday 7:00pm). Plan and write down how long the major steps will take. You must have a contingency plan if the upgrade is not completed in the alloted time. The contingency plan will be activated if you don't complete all steps by a certain decision point (say 5:00pm Sunday). The contingency plan is where you backout your changes. All work should be completed by the end of change window (7:00pm).
3. You should have a backup installation that can be reinstalled in time between your decision point (500pm SUN) and the end of change window (7:00 SUN). This backup makes up your contingency plan and backout. If you have hardware available, an ideal way to do this is to have separate hardware old vs. new. Then, it is just plug the old installation back in and reinstall is very quick. If no extra hardware, you will need a larger amount of time for your contingency plan.
4. There should be a test plan to verify a good install. Add dummy patients, schedule dummy patient appointment, read your imported items, and create a dummy encounter. This testing all needs to be done before the decision point or the contingency plan should be put in place.
5. I recommend doing the installation a Friday night or Saturday. This gives you a long window of time to complete your work and do the followup testing.
although too late for those having tech support issues now, if basic change control procedures like these are followed there should be reason where the program's crashing leads to a patient care disaster.
Good advice in general, but not always realistic to achieve in the Real World. For those of us getting tech support to do the install we are at the mercy of their schedule, in which there are very limited slots for getting installs done in the near future.
In my case, I arranged for the install to be done the morning of a day off (Monday) and made backups of Amazing Charts Version 3 databases and saved the databases to a few different flash drives, desktops and laptops. Version 3 and the databases were installed on a backup Tablet PC, to be used in case of disaster.
I chose to have tech support do the installation in case there were any known issues to avoid with following the default install routine.
After the install I tested the scheduling and reviewed previous notes/imported items, all of which functioned well. I also tried making a note, writing prescriptions and saving imported items for a fake patient. I did not try printing a consult letter or writing + printing a patient instruction letter. I also failed to check whether or not the program was installed under my receptionist's profile in the reception computer (the tech installed it under my profile as "administrator" and I assumed the tech had made the install available to all users who access the computer). I should have known better than to skip testing ALL parts of the program after the install, but I assumed everything was okay.
When disaster struck on Tuesday morning we had already started clinic, so it was too late to roll back the reception computer to Version 3. Using my tablet with Version 3 was not a realistic option, since there was no server (reception desktop) to obtain new patient charts from. Paper was the obvious solution while we waited (in vain) all morning for tech support to return our calls. Lunchtime was spent troubleshooting and eventually a few DLLs were reloaded, stopping the program's continual attempts to reinstall itself. On Tuedsay afternoon, crashes started occurring with writing the patient instruction letters, but we held off rolling back to Version 3 expecting (naively) that tech support was going to call back to address the issues. For the next 3 days we kept calling tech support and waiting by the phone like a lonely teenage girl: "Why won't [he] tech support call me?"
As much as I like Amazing Charts, this episode has taught me that I need to assume that things will go wrong and to be better prepared if they do. I could have avoided all of these headaches by simply keeping an extra laptop with Amazing Charts 3 loaded on it, ready to take over as the server. Live and learn...
