How Much Is that EMR in the Window?

What really are the actual first-year costs of implementing a state-of-the-art EMR in a small group practice?

It’s like comparing Chevrolets to Porsches. They both get the job done, but one has extra features for a large difference in price. Of course, I’m referring to the recent article in Health Affairsthat estimated the cost of implementing an EMR in small (2-12 people) primary care groups. The authors concluded that the first year fixed and “non-economic” costs came to $46,300 per physician. This was a well-designed study to track actual costs, as well as calculate personnel costs. Fleming, et al. estimated the time spent implementing the system by various personnel and multiplied those hours by the hourly wage of those workers to arrive at the “non-economic” costs. Unfortunately, this figure may become the new standard of what physicians should expect their EMR implementation to cost over the first year. Yet the study results are strictly limited to the particular circumstances of the HealthTexas Provider Network. Any attempt to generalize their results would result in gross inaccuracies. As a result, the article may be more misleading than informative.

The article gives a passing mention that the results generated “may mean that our results are not generalizable to small, isolated practices or other practice settings.” I will expand on this understatement.

First, personnel costs included several layers of personnel that are simply not present in most practices.

Labor costs: referred to as Non-Economic Costs, included:

· Network implementation team: $28,000 per practice (> $7,000 per physician). This expense is unnecessary outside of a large health care system.

· Practice implementation team: About $8,000 per practice

· Practice end users (ie, physicians and clinical staff): about $10,000 per physician. The largest part of this figure comes from the estimated 42.5 hours of physician time entering information from paper charts (at $150.00/hr). Why physicians would even be involved in this part of paper to electronic transition is not clear.

Therefore, the combination of training and implementation was calculated to be about $19,000 per physician. These are not actual costs, however. Since the people doing all of this are already employed by the HealthTexas Provider Network, these figures do not represent additional expenditures.

Hardware costs: $25,000 per physician for switches, cables, and wireless Internet connections. Seems like a lot of money for items that aren’t very expensive. In addition, it was calculated that expenses included $7,000 per physician for computers, printers, and scanners. Most practices, even without a current EMR, already have the majority of these components. It is puzzling how this money was expended.

Software and Maintenance costs:estimated at $17,000 per physician per year. How much is attributable to GE Centricity and how much was due to the additional costs incurred by licensing with the two other entities described is not listed. HealthTexas Provider Network felt it necessary to augment the GE Centricity software capabilities with two other products—Clinical Content Consultants’ advanced forms and clinical messaging and DocuTrack from Integra. The latter two additional content providers are not necessary for small practice EMR implementation, but how much cost they add to the HealthTexas Provider Network calculated figure is not provided.

I think it is significant to note that an article that purports to provide an exact calculation of the costs associated with the implementation of an EMR did not evaluate any potential practice savings realized by that implementation. Perhaps the personnel costs incurred by EMR implementation are more than balanced out by the savings realized by eliminating chart pulls, automated patient reminder calls, and improved workflow efficiencies (eg, electronic prescribing and test ordering). It would certainly have been interesting to see differences in physician productivity as well as the practice profitability before and after EMR implementation.

I have reviewed the pricing structures of several of the most popular commercially available EMRs. I assumed a SaaS model in each of these determinations. This means the EMR software and data is housed at a distant site and available with any Internet connection (apparently the model of GE Centricity used by the HealthTexas Provider Network). For both the EMR and practice management software packages, the average cost of both training and Implementation is about $8,000 per physician. This figure does not include the “non-economic” costs that were included in the HealthTexas Provider Network study. The average monthly fee for licenses, maintenance, and support is approximately $6,00.00. This would render an annual cost of $7,200. Therefore, the average actual cost per physician to get a state-of-the-art EMR and PM system (qualifying for meaningful use and ARRA stimulus funds) is actually about $15,000 the first year. This is approximately one-third the quoted figure of $46,659 as the calculated figure derived from the study. Of course, those practices without the requisite hardware would incur those specific costs—unlikely to exceed $5,000 as a worst-case scenario.

Therefore, for the vast majority of practicing neurologists (as previously noted are in solo or small group practices), the conclusions of this study simply do not apply. It would indeed be a disservice if this study becomes the standard quoted cost of EMR implementation. Perhaps further studies need to be performed in a more representative practice environment.