Capitol Health Associates, LLC

Report
On
The Vermont Blueprint for Health
2007 Information Technology Budget Request

Health Care Reform Commission
Senator James Leddy – Representative John Tracy
Co-Chairman

Blueprint for Health Study Findings

On February 21 of 2006 the Vermont State Legislature requested Capitol Health Associates (CHA) conduct a review and make recommendations on the Blueprint budget request. A sizable portion of the request is slated to fund technology in support of a statewide Chronic Care Initiative. Over the past month CHA has met with the Leadership of the Blueprint and attended a joint executive committee meeting of Vermont Information Technology Leaders and the Blueprint in order to collect information concerning the programs technology needs. We have conducted meetings with individuals involved with both projects and have reviewed the RFP responses of the vendors to assist in determining a fact based budget that reflects the needs of the Blueprint for fiscal year 2007.

Currently a number of assumptions exist making it difficult to develop recommendations, however recent activities in vendor selection have made it somewhat easier to grasp the needs of the project and produce a more evidence based approach to building a budget.   In summary, the new budget numbers contained in the table below submitted by the Blueprint reflect a reduction from the original budget request of approximately $500,000 which reflects the availability of facts and changes in overall assumptions for deployment of an information system.  The legislature may want to also examine project staffing levels and legal fees contained in the Blueprint budget request. As an award to Wellogic IBM would present considerable risks to the project CHA recommends a dual contract award to GE/IDX – Orion.

Overview of Findings

During the process of discovery we found the original Blueprint technology budget to be primarily based on assumptions as opposed to hard numbers, and derived from the belief that:

• Commercial Off the Shelf (COTS) Technology to support the project did not exist
• The Chronic Care Information System (CCIS) would be for the most part a design build project
• The current registry supporting the project could be further developed to support the statewide initiative

Another contributing factor to the inability to develop a fact based budget was a request from the Legislature that VITL and the Blueprint ascertain the feasibility of the two entities utilization of a joint technology solution for their respective projects.    

During the past three months progress has been made on the selection of a vendor to support the two projects. The original RFP fielded by VITL for the medication history pilot was called back and reissued to allow for the inclusion of specifications to support the Blueprint Chronic Care Initiative (CCI). At that point 4 vendors remained in the selection process; GE/IDX, Orion, Wellogic, and Quovadx, each revised their offerings to include the additional specifications which included a pricing structure from which the Blueprint could determine a more targeted set of numbers.

Though there was a basis to construct a more factual based budget, it was still difficult to develop a hard budget because of widely varying numbers and some “to be determined” items in proposals from the four vendors. For example, Orion’s RFP response itself included assumptions in several important and potentially expensive areas of their proposal that could cause the Blueprint to have a shortfall on their budget projections by, in our opinion, as much as 35%. These assumptions included the addition of 3M as a teaming partner on the project, offering their vocabulary engine as a solution to a particular shortcoming of the Orion proposal.

Another example is 3M, who offered a pilot price for the technology which was substantial, but failed to include a price for a full rollout of the technology. Additionally, Orion failed to present a hosted solution in their proposal leaving the respective projects at risk and open to a potentially expensive and untested solution. As of 3/24/06 further advancement in the selection process has produced a more defined set of numbers through a series of negotiations between Greg Farnum and the vendors.

This realization narrows variables in possible budget numbers for deployment, but fails to present a concise set of facts from which to work. The need for the 3M engine contained in the Orion response has likely been dismissed as either GE/IDX or Wellogic posses the technology needed to meet the projects requirements. Issues of hosting the CCIS solution still remain as Orion has yet to offer a reasonable solution, however there remains a possibility that GE will agree to take on that task.

Of note, a significant issue has been solved. Through a cooperative effort between the Blueprint and VITL the selection process has been opened to the possibility of a duel award allowing the selection of best of breed for the particular requirements and integrating the solutions into a virtual single platform.

That stated, CHA feels although it is possible to obtain a number closer to actual dollars needed for deployment, many assumptions in actual deployment costs still exist

• The actual number of interfaces to clinics electronic medical records (EMRs) remains undetermined
• The actual number of communities to be rolled out in FY07 as well as an overall timeline remains undetermined
• The final vendor selection has yet to be made
• The hosting facility for the CCIS has yet to be determined

Blueprint Budget

Below are the most recent budget projections from the Blueprint. In response I have submitted several questions which are listed below the table.

Questions and Answers: Below are questions CHA asked of the Blueprint after receiving the revised budget.

Q: The budget calls for $400k in Configuration and Deployment Services; I assume these are contracted services from the vendor for requirements gathering, documentation, any system personalization and installation. Am I correct?
A: yes; that is the best estimate of the approach at this time. 

Q: I am also assuming that the $400k for Electronic Medical Record Interfaces includes engineers and specialists for the requirements gathering, documentation and development of the interfaces and field engineers for the install.
A: estimated $20K/interface is taken from the RFP responses - not verified with any actual negotiation with vendor

Q: In reference to the added four Vermont Department of Health employees and the consultant can you please let me know the duties of the two requested Requirements/Implementation Analysts as opposed to the contracted implementation services described above. Please describe why their functions can not be carried out by the program manager in conjunction with the contractors, consultant, and the Public Health Physician - Health Informaticist
A: The analysts would be more technically oriented to convey requirements to vendor engineers - there were staffing recommendations spelled out in vendor quotes that identified both project management and tech staff needed.  If we are to host the applications, we would need network/operations staff to support the environment - the tech positions would be the resource for that need.

Q: Do you have information on current staffing levels supporting the Information Technology portion of the project as of today? Some information is included in your last budget request but I am having trouble identifying the actual number of staff dedicated to IT efforts.
A: 50% of my time; 20% Eileen Underwood; 80% Peter Marsh; 80% Kim Jones.  These resources are mostly dedicated to strategic planning and negotiations with VITL and requirements gathering; some support of existing pilot communities and review of existing VHR capabilities. 

Q: Does the line item for Electronic Medical Record Interfaces include interfaces to Pharmacy Benefit Managers, Labs, Payors or other entities or are they strictly for practice based Electronic Medical Records  
A: Strictly EMR; relying on VITL for the data source interfaces.  Although we have not validated the changes in architectural approach in the event we actually disengage the BP registry from the base VITL proposal.  If we use the VITL CDR as the only data repository, we would interface the virtual registry view application with the VITL data repository.  Have not had the opportunity to get technical validation from either vendor as to viability of that approach. 

Q: Software Support and Maintenance is slated at 350k, is this figure for FY 07 or are you banking for 08. Please clarify why, during an implementation phase of a project this level of support and maintenance would be required. Is training included in this line item? Do you expect major utilization of help desk support during the implementation phase?
A: training would be included; also reflects quoted software assurance - new release; post-implementation tech support; vendor on-site expenses, etc.

Q: Application Service Provider – Hosting Registry Application is budgeted at 240k, I understand the current thinking is that if Orion is selected that the Vermont Department of Health may host the solution in its existing data center. In that case would this money be used to buy additional IT equipment, offset additional services costs in the data center, and provide for additional telecommunications services, or is the line item strictly for a contracted solution?
A: place holder for contingency - if VDH-hosted,  would support the added network costs, hardware, OS, layered software licensing; if ASP-hosted, estimated $20K/mo. 

Q: Finally, as far as the Privacy & Security Protocols for Participants Agreements, am I correct this is a legal fee? Do you think there is an opportunity to share this cost and the product produced with other State entities or have other entities had to do the same type of work, is there work product already produced?
A: Consultant & legal costs for policy development, privacy requirements negotiations with various partners (providers, patients, data sources, etc.)

Recommendations

Vendor Selection

The current selection process has at this point boiled down to GE/IDX teaming with Orion and the Wellogic/IBM team. While each offers a reasonable solution the GE/IDX Orion combination is a stronger offering. Wellogic’s solution to support chronic care is currently in the development stage. Wellogic reports, that it believes, it can have a product available in a 6 month time frame. In my experience with complex systems development, I can report that time frames are rarely met. The fact that the company has presented to the Blueprint only representative screen shots in PowerPoint as opposed to a working prototype and there has been no formal pricing structure for the modules raises a number of risks to the project. If even a portion of the product was in a testing stage I would feel much more confident that the rather tight time schedules for deployment of the Blueprint CCIS could be met. With these facts present I believe the risks of choosing Wellogic over GE/IDX Orion to be formidable.  

Budget Issues: Technical Budget

The new budget numbers contained in the above table from the Blueprint reflect a reduction from the original budget request of approximately $500,000 which reflects the availability of facts and changes in overall assumptions for deployment of an information system.

CHA recommends that the Legislature fund the Blueprint for the Technology portion of the budget at 2.70 million dollars. Our reasoning is slightly different from the Blueprint budget request as stated in the table above for the following reasons:

As it is CHA’s recommendation that VITL accept and award to GE/IDX - Orion team we are basing our funding recommendation on that assumption. The Orion proposal offers two sets of pricing for the Blueprint CCIS one set of numbers for a technology solution for the pilot at $550,000 and one set of numbers for a full implementation at $1,349,564. The project will require an additional $1,325,000 for interfaces, VITL, Hosting contingent and Grants to Providers which carries a total of $2,674,564.

We feel that given the situation that exists today the Blueprint and VITL are in a good bargaining position given the competitive nature of the RFP. This position is not likely to be as strong in FY08 as it is today having fostered a sizeable discount from both GE/IDX and Orion. Therefore we do not recommend accepting the Pilot pricing structure although the numbers are attractive it leaves the programs at risk. We recommend purchasing technology and services for a full implementation of the CCIS platform.

Budget Issues: Staffing Issues

CHA is at a disadvantage in making concrete recommendations to the Legislature on this subject due to a lack of total understanding of the internal workings and business processes of the Vermont Department of Health. However from a business perspective we feel that for the particular phase of this project and the estimated size of the rollout to the communities the project seems to be overstaffed. In the private sector based on the requirements of the project, current staffing levels, contracted services from vendors we would eliminate the two requested Requirements/Implementation Analysts positions. Funding would be better spent in increasing the salary level of the Project Manager and combining the efforts of the Consultant, Public Health Physician, Contractors, VITL, Hosting facility, and existing staff to manage the initial roll out. We would then evaluate the project staffing requirements to support an extended rollout to additional communities for the 08 budget cycle and fund accordingly. 

Budget Issues: Legal Issue

The Blueprint has requested $350,000 for Consultant & legal costs for policy development, privacy requirements negotiations with various partners (providers, patients, data sources, etc.). CHA, while unqualified to make specific recommendations, believes that certain portions of this work have already been accomplished or are well under way. It is our recommendation that this issue be looked at closely to avoid duplication of effort and expense.

Program Advancements

The Blueprint with assistance from its members, its own consultant HLN Consulting, and CHA, has come to a number of conclusions concerning an enterprise technology platform capable of a statewide chronic care system deployment.

• COTS technology does exist to support the project
• A single or combination of vendors engaged in the RFP can fully support the solution
• The existing registry should be retired
• There is no need to do a design build project
• It will be possible to work with VITL on several joint aspects of the project
• There may be an opportunity to do a combined technology project for population risk stratification with Vermont Department of Health Access.
• That deployment of a fully implemented CCIS to two existing communities and one additional community is, though difficult, possible by July 07.     

The Blueprint has made reasonable forward progress in conjunction with VITL to establish the basis for an enterprise wide implementation of a joint technology solution. It is apparent that the two organizations are working out their issues in an open forum and continue to identify requirements that will insure the success of both initiatives.