EMR Efficiency in Internal Medicine: How User Friendliness Affects Financial Outcomes
Lehigh Valley Physician Group – Hazleton Administration, Lehigh Valley Health Network
This study aims to identify EMR inefficiency in terms of clicks to achieve the same inputs that would be in compliance with Meaningful Use standards. Each system varies in the amount of actions and clicks that must be performed in order to achieve the same ends in regards to these standards. As these standards become increasing more arduous over time because of changing phases, it is crucial to utilize the best system available network wide to ensure network wide compliance and communications between practices. The variance between using two different EMR systems requires more resources to maintain, but poses an opportunity to study the efficiency in which the system achieves Meaningful Use Compliance and Outline an efficiency measure so that a system may be chosen to implement network wide. A qualitative analysis with some quantitative elements was performed at LVPG 3080 Hamilton Blvd. and Internal Medicine at the Dessen Center in Hazleton with significant variation in workflow and mouse clicks. The clicks taken over a month at these two practices were extrapolated to fit into the last 4 months in which insurance denial data was taken. Denial codes were reviewed to ensure that monies accounted for in denials were recoverable, while monies already paid or unreachable were excluded from the data set. Times were recorded for patient encounters in an hour block of time to measure check-in, rooming, doctor’s encounter, and checkout to see how the Epic system would work in place of Meditech in regards to the efficiency of meeting Meaningful Use.
*EMR, Meaningful Use, Meditech system, Epic system, Clicks, Provider, Patient Staggering
As of 2009, the government attempted to revolutionize health care by incentivizing the use of EMR systems in hospitals around the country. These electronic medical records systems intend to increase efficiency of the service in terms of waiting times, expedite information transfer and improve overall patient care. However to insure that these programs are utilized in a productive manner, the Office of the National Coordinator for Health Information Technologies implemented terms for EMR systems in the form “Meaningful Use”. Meaningful Use acts as a set of guidelines that an EMR needs to follow in order to continuously keep receiving federal funding and eventually negate potential deductions in revenue due to penalties for not reaching meaningful use measures. These guidelines expand in complexity by phases to promote even more growth and sophistication within the system. The end result of this process is for the any practice participating in the program to come up to today’s technological standards (“2014 Definition,”2014)
Participating EMRs all try to adhere to Meaningful Use but do this with differing amounts of success when considered from a financial perspective in terms of the insurance denials they produce annually (Ciotti, V., & Alcaro, B. 2012). These denials can be caused by a variety of reasons that could be analyzed in the value stream model which is based on LEAN methodology. This methodology emphasizes the elimination waste while the model simplistically gives a macro-level perspective toward achieving a goal (Martin K. & Osterling M. 2014). Lehigh Valley Health Network aims to reduce these denial rates by identifying problems in user interface of the EMR system to lower overall denial rates.
This study focuses on the differences between interfaces in both Meditech and Epic systems which are employed within the Lehigh Valley Health Network. Both interfaces’ user friendliness were evaluated in terms of the number of clicks the provider (either nurse or physician) it takes to complete an established patient’s encounter including check-in, rooming, doctor’s encounter, and check-out.
This study primarily focused on the established patient base of the population to maintain the reliability of the data’s populations. New patients for the current month at that time were excluded from the study but may be included if they were seen again in the following months. Keeping this data pool in mind, understanding the correlation between the clicks of the mouse needed in input areas and the impact rates on denials is crucial to impacting financial outcomes on insurance claims involving demographics collection, patient care and practice efficiency ( Maxwell, M. 1999).
This study utilized a combination of quantitative and qualitative analysis for the Administrative department of LVHN- Hazleton. The study focused on the number of times a person had to input data by clicking their mouse, or clicking their mouse and then typing (each counting as an individual click), as well as the amount of time performing each step in patient care process. These steps included: checking in at the front desk, nurse’s rooming of the patient, the patient’s encounter with the physician, and checking out at the front desk. All participants in the study were selected because of their established patient status to ensure a more consistent amount of clicks in each EMR system, and reflect a larger majority of patients who choose LVPG-Hazleton and LVPG Internal Medicine.
The Meditech system version 5.66 of LVPG- Hazleton and the Epic system at LVPG Internal Medicine 3080 Hamilton Street were used to collect demographics, record patient notes, and schedule future appointments. Health care providers also had the option of using speech recognition technology called Dolbey Fusion version speech magic 7 to produce patient notes, however this didn’t influence of results because the functionality takes the same number of clicks.
All mouse clicks across patient encounters within a given hour were averaged together and then extrapolated to apply to the entirely of a normal business week and then to an entire month’s worth of clicks at LVPG-H and LVPG 3080 Hamilton Blvd. A distribution of time across these four phases was recorded, and weighted averages were calculated because the observation periods varied in the assigned hour block of time for each encounter observed.
Calculations were made to compensate for Hamilton Internal Medicine’s non-staggered patient booking in regards to clicks so that when the Epic system clicks were applied to Hazleton’s Internal Medicine business hours, the staggered approach to patient booking was accounted for. (Fig. 1)These clicks were then extrapolated to fit into the business hours of their respective offices and then applied to the months since the roll out of Epic on February 18th 2015 at LVHN in Allentown to obtain a measure of how efficient outcomes may be received in terms of these inputs over time.
Additional Calculations were made to evaluate the roll out of the Epic system and the Meditech system by accounting for denials that may be recouped to show a system’s potential for improvement. However, insurance denials that have already been paid or produce a result in which no more money may be collected were eliminated from the study. These CPT codes included: 18 (Duplicates), 24 (Charges covered under capitation agreement), 45 (exceeds cap according to pay schedule), 97 (was already paid for in a similar service) and 246 (test run).
Data consisting of established patient denials from the third and fourth quarter of the 2015 fiscal year was taken from the LVPG Internal Medicine at 3080 Hamilton Street, the Administrative Department of LVPG- Hazleton, and the Revenue Cycle Operations Department of LVHN. Clicks were obtained from LVPG Internal Medicine at 3080 Hamilton Blvd. and LVPG Internal Medicine at the Dessen Center, Hazleton.
The data from 3080 Hamilton Blvd amounted to utilizing 28160 clicks per month in its 99212-99215 codes. Since the Epic system’s rollout on February 18th, a provider would expect to click the mouse at least 112,640 times by May 31st. The office billed 3509 patients and received 170 insurance denials with exclusions applied. In this time, the providers have lost 4.84% of insurance charges that may still be collectable. These charges would amount to a grand total of $496,475 and $24,381.84 could be regained from denials. This results in 4.91% of revenues that may still be recoverable from that fiscal period. Without exclusions, the office had a insurance denial rate of 16.06%.
More data from LVPG Hazleton showed that on average a provider seeing 99212-99215codes would expect to click the mouse 52,645 times and 210,580 times in the same amount time allotted in since Epic’s rollout at LVHN. The Dessen Center Office of Internal Medicine billed 3586 patients and received 70 insurance denials with exclusions applied. In the given time, providers lost 1.95% of denials that may correctable. From these billings, the office charged $429,950 and only $8,260 was recoverable. This indicates that 1.92% of revenues are retrievable from the observed fiscal period.
When Epic system clicks were applied to Hazleton business hours and patient booking schedule, the clicks a physician should expect to click the mouse 28,160 times per month. If a provider at Dessen Center Internal Medicine had Epic, he or she would have had clicked the mouse at least 130,624 times to achieve the same amount of billing at that office. The amount recoverable from that office could be assumed to be around 4.61% if it takes the same trend of recoverable data from Hamilton Blvd and would make generate $408,209 in it’s first few months.
This study was limited primarily by time, which warranted the use of extrapolation of clicks to apply to the last five months as opposed to collecting five months of real world data to show clicks over time compared to it’s financial outcomes which did utilize real world data. If this were the case, than a Pearson Coefficient test would have definitively proved the correlation of an EMR’s efficiency in terms of clicks and it’s financial outcomes. If this study’s methods were to be carried out periodically throughout the year, then another tool to evaluate the Epic’s inaugural year at LVHN has been identified.
Data availability was also a concern due to the limited time of the scholar program. Coordination with the appropriate offices was crucial for the completion of the research and involved more than 2 weeks of waiting for data retrieval. Once communications were established, however progress continued as planned. The data itself was also influenced by differing patient scheduling plans. Internal Medicine 3080 Hamilton Blvd. didn’t stagger patients meaning that it had the patient go through the entire encounter before starting another one and they saw only 2 per hour. Internal Medicine at the Dessen Center in Hazleton staggers patients meaning that providers each try to complete their tasks within a 15 minute cycle and complete their obligations for each new patient within that time frame in regards to check-in, rooming, doctor’s encounter, and check out. This never ending stream of work needs to have a faster interface in order to spend less time on entering data and have a more patient oriented approach to care.
The two systems varying financial outcomes could be attributed to the maturity of the systems, and how acquainted both clinical and clerical staff are with it. The results of the clicks of on Epic would be assumed to produce an even lower percentage of insurance denials because it uses fewer inputs to accomplish the same routine tasks during check-in, rooming, doctor’s encounters and check-outs. However, ignoring the systems ages, Epic clearly accomplishes these tasks more efficiently because Meditech uses about 1.86 times more clicks.
From the observed times during encounters it was clear that work flow at Meditech can’t drastically increase work flow in the office because it’s users must put in so many more inputs than their counterparts on the Epic system without streamlining the interface. (Table 3) the streamlining of the system is evident in the amount of additional time the doctor in Epic has to spend with his patient as compared the doctor in Meditech. The epic system makes work flow a breeze with check in and check out both taking around a minute or two with proper documentation already in the system.
Another factor contributing to the work flows at each office was the number of providers at each office. In order to saturate the patients that required care, the staggering of patients was required using the Meditech system. However this invariably led to back ups as patients visits lasted longer than the allotted time slots. On the other hand, Epic system’s 11 providers handled roughly half the amount of patients that each of the 4 providers on Meditech did because they had more providers and therefore didn’t necessitate the staggering of patient encounters. This lack of staggering meant that the same providers spent less time and clicks on the entire patient population because the population was spread over a larger amount of providers.
I’d like to thank those who helped me throughout my research experience and who shared their hospitality with me. Thanks to Pamela Langdon and Barbara Biacco who made this opportunity to study possible and for being indispensable mentors for the summer. I’d like to thank the Offices of 3080 Hamilton Blvd. Internal Medicine, Internal Medicine at the Dessen Center in Hazleton, Vine Family Practice and Physiatry at the Health and Wellness Center in Hazleton for being my gracious hosts. A big thank you to Hubert Huang in creating this program. Thanks to Zack Chamberlain for providing many, many spreadsheets of data to sort through. Thank you to Richard Shellenberger and Marta Avila for showing me around to many of the different offices around town.
2014 definition stage 1 meaningful use. (2014). Retrieved 7/06, 2015, from http://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Meaningful_Use.html
Ciotti, V., & Alcaro, B. (2012). Top HIS vendors by revenue. Retrieved 7/06, 2015, from http://www.hispros.com/editor_uploads/documents/2012%20Top%20Vendors.pdf
Martin, K., & Osterling, M. (2014). Value stream management. [ Value Stream Mapping: How to Visualize Work and Align Leadership for Organizational Transformation ] (pp. 1-27). New York: McGraw Hill Professional.
Maxwell, M. (1999). EMR: Successful Productivity Tool for Modern Practice Health Management Technology, 20(9), 7/06-46-47.
Table 1: 3080 Hamilton Blvd. Internal Medicine
Table 2: Internal Medicine at the Dessen Center at Hazleton
Hamilton Blvd. (Epic)
*Grey areas represent data excluded from graphical data and calculations because it is non-recoverable.
Table 3: Recorded Times for Entire Patient Encounter (in Minutes)
*Times are calculated averages from 3 test trials at each interval with the Total time being the sum of the averages.
Table 4: Clicks from each EMR system and Hypothetical Hazleton Epic Calculation
*Clicks recorded are calculated averages from 3 test trials at each office.
Published In/Presented At
Snyder, C., (2015, July 31) EMR Efficiency in Internal Medicine: How User Friendliness Affects Financial Outcomes. Poster presented at LVHN Research Scholar Program Poster Session, Lehigh Valley Health Network, Allentown, PA.
Research Scholars (Acknowledgements and Co-authored Publications), Research Scholars - Posters