Electronic Health Records for Non-owned doctors - Cost modeling

The third entry in my series on the 10 critical aspects of providing electronic health records to non-owned doctors is about modeling the costs of the project.

Based on the informatics literature, the initial implementation cost of an EHR for private practices averages between $40,000-$60,000 per provider and the cost of maintenance averages $5,000-10,000 per provider per year. Using these numbers, the total EHR implementation costs for our 300 non-owned doctors could be $12-$18 million and $1.5-$3 million per year. Of course, this includes total costs paid by the hospital and by the practices. To understand the economics of the project, we need to inventory all the costs included and who pays those costs. Stark safe harbors provide some guidance here, since Stark separates costs into those which can be shared with hospitals and those which must be paid by the providers themselves. Up to 85% of implementation costs excluding office hardware can be funded by the hospital. Hardware and most ongoing costs must be paid by the providers. We must also consider what costs the hospital should absorb for planning, legal and infrastructure to offer EHR services to non-owned doctors. These startup costs are nearly the same for 30 or 300 doctors, so they are not easily computed on a per provider basis.

Initial Costs
1. Startup costs to be funded by the hospital
Planning
Legal costs
Hosting Site hardware and operating system software

2. Practice implementation costs to be shared between the hospital and practices
Software licensing fees
Technical Deployment services and Workflow design services
Project Management costs
Training costs
Interface costs
Data conversion costs

3. Practice implementation costs to be funded entirely by the practice
Hardware local to the practice


Ongoing costs
1. Maintenance costs to be funded by the hospital
Hosting Site staffing and hardware lifecycle maintenance

2. Support costs to be shared per Stark
Help desk
Practice consulting support

3. Support costs to be funded entirely by the practice
Hardware service and support
Network connectivity


Of course, each of these categories and subcategories has its own detailed analysis. The "hydraulics" of our model must take into account the goals of the stakeholders - the hospital has a fixed capital budget and wants to connect as many doctors as possible. Doctors want as much subsidy as possible. Given the hospital contribution of x million, and a doctor's ability to pay of y thousand, we need to compute the subsidy level and number of doctors included in the rollout. To help with this decision we're dividing our budgets for all the categories above into fixed startup costs and marginal costs to add 100 doctors. We're also categorizing all costs as subsidizable or non-subsidizable.

Over the next 90 days, we'll do our best to achieve economies of scale, negotiate appropriate vendor pricing, and document acceptable service levels. Our Governance committees will review the final pricing to ensure we've achieved a balance of hospital costs, practice costs, and service. We'll also refine our cost models by documenting all the costs we experience in our pilots this Spring.

Our internal staff and external collaborators are doing a remarkable job documenting the costs. We'll know soon if it is possible to use the capital budgets that the hospital has available to create an EHR product at a price that clinicians are willing to pay for.