Slide Presentation from the AHRQ 2008 Annual Conference
On September 9, 2008, Dan Mingle, MD, MS, made this presentation at the 2008 Annual Conference.
Select to access the PowerPoint® presentation (1 MB; Plugin Software Help).
Electronic Health Record (EHR) Implementation and Adoption Ambulatory EHR at MaineGeneral Medical Center
Tuesday September 9, 2008
AHRQ Annual Conference
We Have a Successful Ambulatory EHR Implementation In Rural Maine
- Funded by AHRQ
- Health Information Technology (HIT) Implementation Grant Number: UC1 HS15337
- Grant Title: Improving HIT Implementation in a Rural Health System
- September 2004-March 2008
Dan Mingle, MD, MS
- 14 years—Clinical Rural Family Practice
- 5 years—Residency Faculty
- Residency Assistant Medical Director
- 2003 Graduate Dartmouth's Center for Evaluative Clinical Sciences
- IT Project Director
- Principal Investigator (PI), IT Implementation Grant
- Director, Massachusetts General Hospital (MGH) Ambulatory Clinical Informatics
- Healthcare Informatics' Innovator Award, 2008
In my 15 minutes
- I told you we were successful.
- I will Describe:
- Utilization data
- Whirlwind tour of the Return on Investment (ROI) data that proves we failed.
- Reiterate that we were successful
- Fight about it in Q&A
The slide shows both a bar graph and a map of the state of Maine.
- The map of the State of Maine shows a border around the Primary Service Area, where MaineGeneral draws patients for primary care services and basic hospital services. It shows a larger area that is the secondary service area from which we draw patients for specialty care. The map shows the locations of MaineGeneral's 2 inpatient hospital locations in Augusta and Waterville.
- The graph explains, from the top down,
- MaineGeneral is an Integrated Delivery Network
- With 2 acute care hospitals
- Spans 5 rural Maine Counties
- 11,000 Square Miles
- Has 112 medical practices of which 30 are participating in the Electronic Medical Record (EMR)
- Has 256 affiliated physicians of which 100 are participating in the EMR
- Serves 140,000 patients of which 70,000 have been seen at least once in an EMR participating practice in the last 18 months
It's about Sharing
The slide shows an image of a blue, vertical oval with four, red, horizontal ovals intersecting it.
- One Patient; One Chart:
- Primary Care
- Med List
- Allergy List
- Problem List
- Care Plans
- The primary innovation of our project is the shared community record.
- One Patient, One Chart means that each EMR participating physician providing care to any individual patient works from the same med list; allergy list; problem list; repositories of results, notes, and care plans; sees the same care reminders; and shares a secure clinical messaging system. Med lists, allergy lists, problem lists, and notes are available in the Emergency rooms and to the hospital admission processes to aid safe, accurate care and enable medication reconciliation.
- The chart is shared across disparate corporate or business lines. The unifying factor is credentials in the MaineGeneral medical staff.
A Short History of a Long Project
The slide presents a combined line graph and time line.
- 2001: Request for Proposal (RFP)
- 2002: Touchworks
- 2003: 7 Pilot Sites
- 2004-2008: AHRQ Grant
- 2007: 1st Ext Pract
- Line Graph
The vertical axis, cumulative e-practices, goes from 0-45 and the horizontal axis, years, goes from 2001 to 2010. The results show:
- 2001: 0%
- 2002: 0%
- 2003: 0%
- 2004: 5%
- 2005: 7%
- 2006: 16%
- 2007: 29%
- 2008: 41%
- The project began in 2001 with the production and distribution of a Request for Proposals
- In 2002, the Touchworks product from the vendor, Allscripts, was chosen.
- Implementation in 7 pilot sites began in 2003.
- Grant funding was awarded in 2004 and ended in 2008
- Early practices were all owned by MaineGeneral. The first independent practice was added in 2007
Tracking the Conversion
The bar graph presents the results for "TaskUser" from Qtr 1 2003 through Qtr 1 2008.
- The utilization of tasks has proven to be the most reliable indicator of participation in the EMR. This graph shows the steady increase in unique task-users in the system by quarter. In the first quarter of 2008, there were nearly 600 distinct users seen in the system.
The slide shows two bar graphs, one measuring the results for "Problem entries per Provider," from Qtr 1 2003 through Qtr 1 2008, and the other, the results for "Problems per Patient," from Qtr 1 2003 through Qtr 1 2008.
- If there is good reporting access to the clinical database, a good problem list is good disease registry. The graph on the top left shows the increase in average number of problem entries per provider per quarter. There was a large increase in problem entry in the first quarter in 2005 that coincided with a phase 2 rollout of structured notes for visit documentation. The graph shows a fairly rapid stabilization of problem entry volume by the 4th quarter of 2005.
- The graph in the lower right hand corner shows a steady increase in the number of problem entries on each patient chart. No steady state has yet been reached through the first quarter of 2008.
Structured Data is Key to Clinical Reporting
The slide shows a bar graph measuring the results for "Structured Findings per Patient," from Qtr 1 2003 through Qtr 1 2008, and a line graph presenting the results for "Usage Categories by Number of Findings per Patient," from Qtr 4 2004 through Qtr 1 2008.
- Structured data is the only data that is reportable data using electronic query tools.
- The graph in the top left corner shows the appearance of structured history and exam findings on patients beginning in the first quarter 2005, when structured notes were distributed and taught. The graph shows a slow overall increase in utilization of those findings and also shows a cyclical pattern that coincides with the expected July influx of new resident physicians and new physician recruits.
- The graph in the lower right corner shows a steady increase over time of the utilization of structured findings as evidenced by the number of providers using, respectively, less than 3 structured findings per patient, 3-13, 14-25, and over 25.
- I rarely experience errors in the EMR.
- I rarely lose any of my work in the EMR.
- Overall, the speed and reliability of the EMR meets my needs.
- The EMR is about as fast as other MaineGeneral systems.
- The EMR support team and Web site is there when I need it.
- Note: The bar graph measures the percentages of "Strongly Agree," "Somewhat Agree," "Not Sure," "Somewhat Disagree," or "Strongly Disagree" for the questions above from 2006 to 2008.
- A questionaire was added to the login screen in 2006. 5 questions are posed on the questionaire which prompts one of 5 graded responses and space for textual clarification. One summary graph is provided aggregating the balance of responses to all 5 questions by year. A steady improvement in user satisfaction is noted
It's Not About EMR
The slide shows a circle being dissected by two arrows, one vertically and one horizontally. The top arrow points to four intersecting ovals: "Needs," "Individual," "Wants," and "Community;" the right arrow points to "Access;" the left arrow points to "Quality and Safety;" and the bottom arrow points to "Cost."
- It's About A Better Value for the Healthcare Dollar
- The reasons that MaineGeneral implemented ambulatory EMR is summarized in this slide. The value compass shows that MaineGeneral seeks to improve delivery of care to individuals and overall care to the community, addressing both the "wants" with which patients present to the office, and the "needs" that they may not know that they require. MaineGeneral seeks improvement in both quality and safety of care and of access to care. And MaineGeneral seeks to improve, from the practice perspective, the cost of providing care and revenue derived from care and, from the patient's perspective, to reduce or contain the rise of overall healthcare costs.
Mixed Quality Measures
The slide shows a bar graph measuring both the "Baseline" and "Final" percentages for the following:
- A1c Tested
- A1c less than 7
- A1c greater than 9
- Lipids Tested
- Low density lipoprotein (LDL) less than 100
- Microalbumin Tested
- Nephropathy on angiotensin-II converting enzyme (ACE)
- Dilated Eye Exam
- Foot Exam
- Flu Vaccine
- Pap Screen
- Colorectal Cancer Screen
- Using chart audit methodology before, during, and after implementation, a mixed pattern of improvement and regression is seen.
- A1c and LDL are both tested a little less often
- But A1c and LDL control are both better overall.
- Microalbumin is more often tested, but ACE inhibitors used in proteinuria a little less often.
- Dilated eye exams are more frequently documented as are foot exams.
- Flu shots and pneumovax are less frequently documented.
- Mammograms and Pap smears more often documented.
- Colon cancer screening a little less often performed.
Near 100% ePrescribing
- 25,000 e-Prescriptions per Month:
- 1,400 Schedule II prescriptions
- 2,400 Schedule III to V
- The line graph presents the results for "C2," "C3-5," and "Total." The vertical axis goes from 0 to 30,000 and the horizontal axis goes from 2/1/2003 to 10/1/2007. Both C2 and C3-5 remain relatively flat until 4/1/2005 where a small increase occurs. Both end on 10/1/2007, between 2,000 and 3,000. Total, however, steadily increases reaching 24,000 by 10/1/2007.
- ePrescribing was a quick win across the organization. There has been near-100% utilization of electronic systems for prescribing. MaineGeneral generates over 25,000 electronic prescriptions per month.
The slide shows a bar graph measuring the percentages of "Voice Mail," "Phoned In," "Handwritten," and "Electronic" error rates.
- Voice Mail: 22%
- Phoned In: 5%
- Handwritten: 4%
- Electronic: 2%
- A count of callbacks to the doctors office performed at one local pharmacy yielded comparative error rates for each of 4 prescribing practices that effectively demonstrate one aspect of improvement in prescription safety growing out of eprescribing.
The slide shows a bar graph measuring the results for "Enterprise Patient Panel Size" from Qtr 1 2003 through Qtr 1 2008. The results show a steady increase from 25,000 in Qtr 2 2003 to 58,000 by Qtr 1 2008.
- Panel size has grown across the EMR participating groups. But the growth has been consistent with the addition of new practices and has not tracked to increases in patient panel sizes attributable to individual physicians.
The slide shows two bar graphs, one measuring "Visits per Patients per Year" from Qtr 4 2003 through Qtr 1 2008, and the other "MGHA PCP [primary care provider] Visits per Provider Hour" from Qtr 3 2002 through Qtr 1 2008.
- The graph on the top left shows an overall decrease in the number of visits per patient per year over the course of the project. This graph shows some cyclical changes not readily attributable to EMR.
- The graph on the lower right shows an overall decrease in the number of visits per provider hour. This matches the report of many physicians that they "see fewer patients but generate more complex E&M codes.
Revenue and Expense
The slide shows two bar graphs, one measuring "MGHA PCP Revenue/Expense per Visit" from Qtr 3 2002 through Qtr 1 2008 and the other, "MGHA PCP Total Revenue and Expenses" from Qtr 3 2002 through Qtr 1 2008; and one line graph, which presents the results for "MGHA PCP Gain (loss) per Visit" from
Qtr 3 2002 through Qtr 1 2008.
- These graphs document changes in the revenue and expense relating to patient visits. All graphs document a phenomenon commonly reported in hospital—owned primary care practices, that the cost of providing care exceeds the revenue generated.
- The top left graph shows cost and revenue per patient visit documenting a balanced increase over time.
- The graph on the top right shows total revenue and expenses, also documenting an increase in both over time.
- And the graph on the lower right shows only random variation in the net loss per patient visit over 5 years of the project.
Little Difference in Hospitalization
The slide shows two line graphs presenting the results for "Inpatient Stays per 1,000 Patients" from 2003 through 2006, and the other, "Inpatient Costs per Patient" from 2003 through 2006.
- Hospitalization rates and costs were obtained from the Maine Health Information Center all claims database and compared EMR primary care practices to other Maine primary care practices. The number of hospitalizations parallel each other fairly closely. The cost of hospital care for EMR practices suggests a selective increase in the cost per hospitalization in 2006 in MaineGeneral hospitals.
The slide shows a duplicate copy of the circle being dissected by two arrows, one vertically and one horizontally. The top arrow points to four intersecting ovals: "Needs," "Individual," "Wants," and "Community;" the right arrow points to "Access;" the left arrow points to "Quality and Safety;" and the bottom arrow points to "Cost."
- Equivocal improvements in Quality
- Significant improvement in Safety
- No change in Access
- No attributable changes in cost or revenue
- Dr. Perspective
- Pt. Perspective
- This slide summarizes the data presented in the last several slides to the value compass previously introduced.
- 30 practices
- 100 Doctors
- 30 Residents and Fellows
- 12 practices in the implementation queue
- 70 more to go
- For 47,000 patients
- To 550 users
- 17/patient; 1500/user
- For 48,000 patients
- With 190 Providers
- 4/patient; 960/provider
- For 36,000 Patients
- By 170 Providers
- 6 /patient; 1900 / provider
- 1344 EMR Help tickets
- This slide summarizes the utilization and acceptance of ambulatory EMR in the MaineGeneral region, documenting vigorous utilization in 2007 and an implementation queue carrying the project beyond the period of the AHRQ grant.
- By utilization and sustainability criteria, the project might be considered a success.
- By ROI measures, though there are clear safety improvements, quality improvements are equivocal, there is no significant change in patient access and no change in cost of providing care, cost of receiving care, or revenue generated from care. From the financial standpoint, the project has not yet shown success, unless the lack of significant loss of productivity can be considered success.
Current as of January 2009
EHR Implementation and Adoption Ambulatory EHR at MaineGeneral Medical Center. Slide Presentation from the AHRQ 2008 Annual Conference (Text Version). January 2009. Agency
for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/about/annualmtg08/090908slides/Mingle.htm