AnsweredAssumed Answered

Designing an iPad-based anesthesia record

Question asked by dsimonson on Jun 6, 2015


Designing an iPad-based anesthesia record


Ok, I have looked over anchor-buoy, selector-connector, session models, data separation models, and many other ideas.  I have my initial plan ready.  Please see my uploaded “Manage Database - Relationships” screen shot, I would love to hear your opinions.  Given the length and complexity of this post, I’m not sure this forum is appropriate, but constructing it has helped me organize my thoughts, and any comments, even breezy ones, are greatly appreciated.

The scenario.  

I am constructing a solution that a solo or small-practice Certified Registered Nurse Anesthetist (CRNAs) can use in a single-user situation on an iPad to provide them with an electronic anesthesia record.  It will not (as the basic design) be connected to a server - it is meant to be used in the field.

My vision is to make it free via some type of appropriate shareware licensing.  Note that it needs to be HIPAA compatible, thus I will have to include audit logging and tracking and so on.  I am going to use FileMaker 14 to develop it.

The workflow.  

The user will create a new patient record (or look up an existing one, although there are usually not more than one or two anesthetics per patient).  They will then fill in the data on the patient - demographics, insurance, History and Physical, allergies, etc.  

Once the data is filled in, they will start the case by identifying the anesthesia technique and monitoring.  They will specify an anesthesia start time (very important for billing the case).  

At that point, they will be taken to a skeuomorphic anesthesia record (by skeuomorphic, it looks like what we currently use on paper).  They will then begin to enter timed “observations” - blood pressure, medications administered, gases administered, heart rate, etc.  These observations will then show up on the anesthesia record - which is really, at this point, a summary report of these observations in graphic form.

At the end of the case, they will finish the record and then print it to pdf and transfer it to the facility to be included in the patient’s chart at the facility.  In addition, a supplies used list will be produced, as well as a properly-formatted xml file (or something like that) for transfer to the billing company.

Now to my structure.

I really like anchor-buoy, so I am definitely including that.  I am not sure that connector will work for me, however - and I’m worried that it’s because I don’t know better.  I have included it in my relationships chart anyway.  For example, my medication observations will consist of fields for observation id, anesthetic id, medication id, amount, and time.  It would be great to have the meds available via to the the “connector”, but in looking at my relationship graph I don’t see it working.  It can work for my pick lists, and then insert the chosen med id into the observation field, but I don’t see how it will work to display the name of the medication (instead of just its id) without adding a TO for meds hanging to the right of my med observations TO - the thing connector is supposed to prevent.

I also don’t see how “selector” will work in my scenario, other than to help populate lists of available medications or whatever.  I will usually only be dealing with one patient at a time, and selections will be from drop-down lists (or popovers) as the observations are populated.  Again - this may be because I just don’t understand the selector concept well enough.

Data Separation model

I will need a robust system for end of day archiving of finished anesthetics which allows transfer off of the iPad.  I see this solution as only keeping one or two day’s worth of anesthesia records at a time, and will encourage users to offload records as they build up.  

You can see that I have a large number of tables in my graph - I think I will plan on having all of the tables containing Protected Health Information (PHI) in one FM file, and the rest separate.  Then, to offload the patient data, I will copy off the PHI file and replace it with an empty clone.  

Ok, enough at this point.  To anyone who has read this far, thank you for your patience - and please blast away!  I am learning so much so quickly that pointing out any errors will really help, even if you don’t read all of this.  Thanks!

Dan Simonson