The following is an extract from a LinkedIn discussion on how to reduce the chances of failed EMR/EHR implementations.
# The perfect electronic health record and care management system
One that retains the best features of ‘paper charts’ (everything relating to the Pt in one place) but gets rid of the bad features (solves the “where’s the chart” problem, adds filtering and sorting, provides decision support without imposing it, supports interoperability)
# The ideal workflow patterns
The ones the agency has spent years building and refining to give the agency a competitive advantage (avoid “solutions” where “ . . . you now need to do it this way because the system needs it to be this way”).
# Data sources, responsibilities and priorities
Data from the outside world (filtered and validated), from internal policy/procedure, from data mining done across different patient populations, from information picked up at steps along best practice protocols, from decision support algorithms;
Responsibilities are a non-issue, in that if best practice protocols have skill attributes defined at steps, auto-resource allocation software will route tasks to those who are supposed to perform, them;
Same solution for priorities – process logic largely determines sequencing, users micro-schedule based on workload and can invoke problem escalation, supervisors are able to level and balance across staff)
# The necessary linkages
Make it easy, use an EMR that takes data from the UI and pushes it both to the patient EMR and to an official Data Exchanger where interested prospective subscribers can go to the owner of the data and subscribe to specific sub-sets (need to know basis).
Make it such that they can pick up their data using subscriber data element naming conventions.
By all means define minimum data sets, but let’s not go to some rigid structure where everyone has to fill in the widest possible number of ‘standard’ data elements to satisfy all possible data recipients). Goes against the MY workflows, MY forms, MY data approach I keep saying is the key to efficiency.
# The greatest deficiencies (ie: money, human resources, training etc.)
Not money (because internal “money” comes out a different bucket from outside consulting not that agencies can or should do everything on their own);
Not human resources because the methodology of documenting/ improving/deploying processes no longer needs to be done on ‘stone tablets’;
Not training because users see THEIR workflows posting THEIR forms to a simple UI consisting of ONE screen, so, what training, really, is needed other than to introduce them to the computer and point out the on button, screen, keyboard and mouse ?