Ever wonder whether it would be smart to drop your vendors and grow your own EMR/HIT apps? Here’s some food for thought.
Today, I ran across a very detailed analysis of the labor costs involved in developing EMR and HIT applications, straight from the blog of the always perceptive Shahid Shah.
Shah, a longtime health IT consultant and former CTO for two EMR companies, just posted a list of the various professionals you’ll need to develop such applications — and the high prices these professionals command:
- Clinicians and healthcare professionals (HCP) like docs, nurses, etc. – perhaps as consultants
- Senior project manager – about $150k per year
- User interaction engineer (UX, usability) – about $120k per year
- Mobile app developers (iOS, Android, etc.) — about $90k per year
- Database modeler and information architect (SQL) — about $150k per year
- Database administrator (SQL) – about $120k per year
- API engineer (REST / SOAP) – about $120k per year
- Service code engineers (Java, Ruby, etc.) – about $150k per year
- Security analyst and privacy engineer (HIPAA, HITECH, Sarbox, etc.) — perhaps as consultants, $175k per year
- Cloud infrastructure admins (Amazon, Eucalyptus) – about $90k per year
- Network infrastructure admin / engineer (TCP/IP, etc.) – about $120k per year
- Data integration engineers (ESB / ETL / connectors) – about $90k per year
- HL7 and healthcare data integration conformance engineers – about $90k per year
- Technical documentation specialist – about $60k per year
- Quality assurance directors (test strategy, test planning) — about $120k per year
- Quality assurance engineers (test planning, manual execution) – about $80k per year
- Quality assurance automation (automated execution) engineers – about $90k per year
- Trainers (folks with healthcare office experience plus tech knowhow) — about $60k per year
As Shah notes, this is what U.S. specialists typically cost. (Working with Indian developers can save you about 35 percent, he estimates.) Still, either way we’re talking about a bundle on compensation alone.
Despite the expense, there are probably some large institutions which will choose to develop EMRs or related applications internally.
After all, if you have a deep enough IT bench, developing even high-end applications might be cheaper than paying for high-end packaged products. And of course, there’s always something to be said for apps developed exactly to your own specs.
Ideally, your institution could build its own EMR/HIT apps, then license them to other institutions or co-develop them with partners who can sell them elsewhere. (For an example of how this might work, check out the $400 million partnership deal the University of Pittsburgh Medical Center did with IBM a few years ago.)
Still, Shah’s analysis is more than a little sobering, isn’t it?