Several physicians contributed to the project, though there was no formal budgeting for their time. The lead informatics physician on the project was allocated 1 day/week to complete various information technology (IT) projects. The lead cardiac anesthesiologist contributed nonclinical time. Approximately 95 physician hours were required during the 12-month build, split disproportionately between the first 2 months (45 hours) and the 8 months of development and testing (50 hours). The proof-of-concept and feasibility stages involved mapping the existing workflow (approximately 20 hours) and the billing process (5 hours), and identifying compliance issues before being able to identify points of breakdown and draft the target design specifications (20 hours). Physician responsibilities during development are discussed below.
This investment was necessary to determine exactly what the project would need to accomplish systematically. The project benefited from physicians with expertise in both the clinical and technical domains so that poor assumptions were not made, which could have had disproportionate downstream effects. The complexity and scale were then assessed by IT to determine the technical feasibility of moving forward. Total IT time was approximately 300 hours (6 hours/week for 49 weeks), with an estimated final project cost of $31,375 (Fig. 5). This time estimate only accounts for resources at Stanford. It does not account for consulting work from the EMR provider team as build challenges were encountered.
Once the new system was in place, 99 of the 117 (84.6%) cases performed between August 14 and October 18, 2014 were billed. From this increase we anticipate full project cost recovery in approximately 18 weeks. It is worth noting that the old billing process was subject to volatility due to the micromanagement required to create a charge. The decreased billing for the months immediately before go-live is attributed to focus being shifted to deploying the new system. This explains why a year-on-year analysis was done rather than comparing against the preceding months. We anticipate this volatility will be reduced by the new system.
User response has been positive. Chart review is easier because the TEE note is immediately available for referral by the other clinical teams. In addition, the departmental expertise gained can be applied to planning or consulting on future projects. The template for the procedure note is available for reference and use at other institutions using the same EMR vendor. Finally, quality control reviews have been facilitated now that all TEEs are indexed in a searchable database.
When proposing a wish-list for improving an EMR, misconceptions of what is possible, feasible, and cost-effective can and will occur. Clinicians with no background in computer science may not realize the complexity of what seems a simple request. Software engineers with no clinical knowledge may not be focused on the clinical utility of their product. We found that including experts from clinical and IT domains provided the flexibility necessary to address issues quickly while maintaining the project’s momentum. It is also important to remember that most project costs occur in the development phase. Money spent up-front to thoroughly define the problem and assess feasibility will limit ballooning costs later.7 Ultimately, such projects should have well-defined returns on investment. Is the goal clinical utility, revenue, knowledge, prestige, or some combination? Explicitly stating the targeted outcomes will allow setting reasonable expectations.
Any new system will be met with some pushback from users. Change is difficult. Successful implementation is aided by early and frequent feedback, an agile design methodology that is effective.7 Iterative prototypes help users of the system quickly and clearly define what they want. To accomplish this, the lead cardiac anesthesiologist demonstrated in-progress versions to colleagues often throughout development so that the final tool was clinically useful. As is often the case, feedback was most forthcoming when done one-on-one. The flexibility to create either a brief, basic report or a detailed, technical one from the same template was highly valued. Users also valued the ability to add free text findings to the mandated template fields. Frequently seeking input kept end-users personally invested in the creation of the new system. It also increased general familiarity with the system, which ultimately enabled a smoother transition.
The complexity of the TEE examination and the extensive use of scripting pushed the EMR to its functional limits. It required significant system resources to load and save the procedure note and associated macro. Template responsiveness remains an unresolved issue. It may also be useful to test the process against a mock audit to discover remaining weaknesses.9 Notably, the system does not force the attending to finalize the archived images as it does with the EMR report. Finally, the question of who should benefit from the derivative value created by distributable projects like these is largely unanswered. The considerable investment of time and resources to develop this procedural note are now available to any Epic licensees. Solutions that encourage development of similar projects rather than incentivizing a wait-and-see attitude remain an open debate.
Although the former system was sufficient to provide good clinical care, the EMR was underutilized despite being able to provide a faster, more efficient, and more consistent workflow. It is not enough that health care providers merely adapt to these new tools. Organizations should have expert clinical users who can identify areas where improvement can be made and lead projects to use EMRs to greater effect. The financial investment alone should prove well justified. Any additional benefits derived from increasing institutional expertise will be useful for future projects and for advising others who are attempting to tackle similar problems. Lastly, while Meaningful Use was not the primary objective, this project added value to the clinical process by bringing another component of care into the queryable database.
The authors wish to acknowledge Zhi Liang and Elizabeth Baptist for their assistance in providing detail for this manuscript.
a CMS Meaningful Use EHR Incentive Programs. Available at: https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/. Accessed October 19, 2014.
1. Samal L, Wright A, Healey MJ, Linder JA, Bates DW. Meaningful use and quality of care. JAMA Intern Med. 2014;174:997–8
2. Cebul RD, Love TE, Jain AK, Hebert CJ. Electronic health records and quality of diabetes care. N Engl J Med. 2011;365:825–33
3. Kadry B, Sanderson IC, Macario A. Challenges that limit meaningful use of health information technology. Curr Opin Anaesthesiol. 2010;23:184–92
4. Duszak R Jr, Nossal M, Schofield L, Picus D. Physician documentation deficiencies in abdominal ultrasound reports: frequency, characteristics, and financial impact. J Am Coll Radiol. 2012;9:403–8
5. Behta M, Friedman G, Manber M, Jordan D. Evaluation of an automated inferencing engine generating ICD-9CM codes for physician and hospital billing. AMIA Annu Symp Proc. 2008:873
6. Fahy BG. Acute pain management efficiency improves with point-of-care handheld electronic billing system. Anesth Analg. 2009;108:583–7
7. Beck K, Andres C. Extreme Programming Explained: Embrace Change. 20042nd ed San Francisco, CA Addison-Wesley
8. Mathew JP, Glas K, Troianos CA, Sears-Rogan P, Savage R, Shanewise J, Kisslo J, Aronson S, Shernan SCouncil for Intraoperative Echocardiography of the American Society of Echocardiography. . ASE/SCA recommendations and guidelines for continuous quality improvement in perioperative echocardiography. Anesth Analg. 2006;103:1416–25
© 2016 International Anesthesia Research Society
9. Schildmeyer T. Billing: the lifeblood of your business–does yours measure up? A roadmap to assess your billing performance and audit your billing practices. J Med Pract Manage. 2010;26:133–8