New forms

drbowen wrote on Tuesday, March 22, 2005:

As you start developing new compatible forms, please do us all a favor and spend a little more time on the printing function.  The current printing (and spelling) sucks.

Medical forms are not just true/false.  They are yes, no and “I didn’t look” (null).  All of the current forms default to “on” or “I didn’t look”.  This makes for a very confusing printed document.  What we send (usually to insurance companies) is a long list of "on"s and “I didn’t look”.

This will not do well in an insurance audit.  Neither are other physician offices impressed with our gramatical accuracy or clarity of documentation.

We still have to send out a lot of printed documentation either on paper, pdf or other electronic format.

For instance, one of the forms that made the select "at least it works" list has several obvious spelling errors that I have been trying to correct for the last 18 months.

Please have a little pride in your work.

Sam Bowen, MD

tekknogenius wrote on Wednesday, March 23, 2005:

So, what do you suggest? What should the default be for check boxes, and/or radio buttons. I noticed the default on for checkboxes is only displayed in the encounter screen, when it is "reported" the value is set to "yes".

drbowen wrote on Wednesday, March 23, 2005:

Better defaults would be yes, no and ‘null’.  Null should actually be inserted into the database as ‘not asked’ for history forms (such as the review of systems) and ‘not examined’ for physical exam sections.  All of these choices will improve the readibility of the forms.

Radio buttons would handle these three choices more efficiently than check boxes. Check boxes by their nature have only yes and no/null. The check boxes are actually what is causing the current problem.

Valid default values will depend on the working habits of the practitioners and should be modifiable by the individual physician.  This could be managed with a "update this form" button allowing the practitioner to choose different defaults.

OpenEMR is still very typing intensive.  This will impact its growth rate and the ultimate number of end users.  Physician’s are generally not typists and a blank form like “form_dictation” is very intimidating.

We need a lot more forms like the (now defunct) ankle injury form, bronchitis form and other chief complaint oriented forms.  A number of high volume clinics are already using a "template" system where specific types of problems are templated with lots of drop down lists. 

I have a lot of clinical data and have previously run a statistically relevant report simply listing the most commonly used diagnoses in our primary care / urgent care office.  Acute bronchitis was the number one diagnosis.

Thanks for your interest.

Sam Bowen, MD

tekknogenius wrote on Wednesday, March 23, 2005:

Actually checkboxes are binary by nature, only being able to have two values. Often times, it is yes or no, but it could be normal/abnormal, 1/0. They should be used when there are only two answers, ever.  Therefore a translation is needed to go from database to "view"/"report". This can be handled in the report.php, but report.php is not used in the encounter summary screen. It just dumps the field name and value from the database. I think there should also be a summary.php for the form, so that it can display correctly in the encounter summary screen.

wpennington wrote on Wednesday, March 23, 2005:

You may want to investigate the Review of Systems form in the OpenEMR for Mandrake 2.7.1.  This form is based on a positive, negative and not asked.  The form was also formatted for reporting in the patient reports section. 

rwjblue wrote on Wednesday, March 23, 2005:

How is the OpenEMR for Mandrake rpm different from the main CVS code?  I understand that there has been a significant amount of testing involved from your firm, but shouldn’t all of the corrections that were made be incorporated into the main CVS repository? 

drbowen wrote on Thursday, March 24, 2005:

Thanks for your question Robert.  I believe many of us are thinking the same thing.

Walt,

I don’t use an rpm based distribution.  I am not convinced that there is enough new information in OpenEMR for Mandrake 2.7.1 for me to make the effort to convert the mdk to a source package and then install on my existing systems just to look at one form.

I know your working methods are different, Walt, but most open source projects release the source packages first and then follow then them with rpms later.

For me, Mandrake is definitely a step backward.

Sam Bowen, MD

tekknogenius wrote on Friday, March 25, 2005:

I think they are the same, except setup stuff.

drroller wrote on Friday, March 25, 2005:

When printing the results of an exam, it may be best if only what you performed and results are on it.

If you have to go court to testify on behalf of an auto accident victim or the like, you don’t want…did not perform or blanks on your form because it opens the door for the other attorney to ask…well Dr…why didn’t you perform…where he goes down every did not perform on your form.

I think "did not perform" is a remnant from the paper form days where forms were not dynamic.