The Future of Family Medicine (FFM) report published in 2004 called for a new, idealized model of care that had 11 key elements:1
Personal medical home,
Team approach to care,
Elimination of barriers to access,
Advanced information systems, including a standardized electronic health record (EHR),
Redesigned, more functional offices,
Care provided in a community context,
Focus on quality and safety,
Enhanced practice finance,
Defined basket of services.
A subsection of the report on financing the new model suggested that if one were to achieve all or most of these elements, one could significantly increase physician income through greater efficiency and productivity.2 Now the idea of the patient-centered medical home (PCMH), which embodies much of what was outlined as the new model, has caught the attention of physician groups, policymakers and payers. Growing recognition of the current crisis in access to primary care in the United States, with all that implies for quality, cost and patient experience of care, makes the PCMH attractive. Many payers, employers, patient advocacy groups and providers are working to develop pilot projects to demonstrate the feasibility and impact of the PCMH. These pilots typically depend upon new payment models and means of evaluation, however. Given that, they strike some physicians as overly complex and onerous in the context of the already stressed, poorly organized and poorly staffed environment that is too typical of primary care in the United States.
While these projects have obvious merit, I want to suggest what may be a simpler approach – one that does not depend on new models of financing and that follows a logical series of steps, each building on the one before to develop the capacity and infrastructure for a high-volume practice capable of sophisticated population care and management. It is really a reprise of the so-called “Task Force 6” report of the FFM,2 but I hope it demystifies the process and shows a pathway that most family physicians can imagine themselves taking. These steps are immediately available to the typical practice, draw on key lessons that have been showcased in Family Practice Management over the past few years, and have been implemented by a number of primary care practices in whole or part.
Step 1: Improve documentation and coding
The most recent all-payer data available (which are admittedly old, being from 2002 and 2003)3 show the average family physician coding established patient visits in the following distribution:
99211, 2.14 percent
99212, 14.89 percent
99213, 66.13 percent
99214, 15.79 percent
99215, 1.04 percent
Even in a Medicare population with its inherent complexity of care, family physicians still code half of their evaluation and management (E/M) established patient services as 99213 and only about a third as 99214.4 It is widely believed that, despite residency training and admonitions thereafter, most family physicians undercode and many underdocument.5 Fixing these problems can be of enormous financial benefit to the practice. Assuming current Medicare payment rates, changing to a distribution where most of the codes are 99214 could yield an additional $30,000 to $75,000 per full-time physician per year (how much depends on where you begin and how close you get to an ideal distribution of codes). See “Coding From the Bottom Up” in the November/December 2008 issue of FPM6 for more on coding guidelines, and see “What improved coding and increased volume can do for practice revenue” for an illustration of the financial impact. If one's overall payer mix is more remunerative than Medicare, then the impact would be even greater. This shift would have to stand up to audit, but it would be a lifeline to resource-starved practices that currently see no escape from “hamster wheel” medicine.7 We have implemented this in my own practice, moving our distribution to 60 percent level 4, and our compliance auditors say we are still occasionally undercoding. To encourage the physicians in my practice to routinely consider whether their services and documentation support a level-4 code, I posted a guide to these requirements at every physician's desk. (A level-4 reference is part of the reading list below.) Drs. Christine and Tom Sinsky in Iowa,8 Dr. Peter Anderson in my home state of Virginia9 and many other Virginia family physicians with whom I have spoken have all made similar improvements in coding, and all of them have passed audit.
Step 2: Hire more nurses or medical assistants
Six to 12 months of coding improvement should get revenue to the point where you can afford to hire enough nurses to raise the nurse-to-doctor ratio to between 1.5 and 1.75. Why load on what looks like more overhead? Because critical analyses of manufacturing and service industry processes show that maximum capacity and efficiency of operation is achieved when you “unload” the rate-limiting step (in most practices, the physician) to maximize throughput, whether you're producing widgets or a service such as patient care.10 The point is to enable the only person “bringing money into the practice” to do what he or she is specifically trained to do – to diagnose and treat patients. The added nursing staff could enable physicians in the practice to offload data gathering, data organizing, chronic disease and preventive services management and data entry tasks. It has been demonstrated that this expands capacity and improves quality.8,9 By having all staff in the practice “working to the top of their license” and by creating highly functional teams, you also create improved job satisfaction, higher employee loyalty and reduced staff turnover.11 We have begun this process in my own practice, and our Fairfax residency program is adopting this model for all of the family physician faculty. As Terry Mills, MD, pointed out in a recent FPM article,12 you have to ask yourself if you would pay another physician to do some of the work you are doing. If the answer is no, then delegate it.
Step 3: Implement advanced-access scheduling
This step, which requires being able to see patients the day they call for an appointment, could be taken at any time, but I suggest that you wait six months after the extra nurses are hired to give physicians and staff some breathing room before taking it on. There is little direct cost to making the transition other than physician and staff time. “Working down the backlog” doesn't have to require adding a lot of new patient care time. For example, you could close the books to future appointments during the slowest month of the year and then re-open the books to advance appointments, but no further out than two weeks. We successfully used this approach in my practice, and although I now have a larger panel of patients and slightly reduced patient care time compared to three years ago, my third-next-available appointment has gone from two months to three to six days. The result is happier patients, happier staff who spend less time on triage and more time on patient care, a lower no-show rate and improved continuity.
Step 4: Increase the number of patients you see per day
The typical family physician sees 20 to 25 patients per day. Increasing this by five per day could bring in an additional $85,000 annually, and increasing by 10 per day could bring in $170,000, assuming optimal coding distribution and Medicare payment rates. It is possible to increase volume once you have completed steps 1 and 2 because at this point, physicians in the practice have offloaded enough “busywork” that they can see more patients per day without adding hours to their workweek. Peter Anderson has shown how adding two specially trained clinical assistants (nurses or MAs) allowed him to go from seeing 25 patients per day to seeing 40 per day without adding hours.9 Increasing volume offers benefits at the societal level as well as the practice level because of the need for greatly increased primary care capacity now and in the foreseeable future.
Step 5: (Optional) Expand hours
If your practice includes two or more physicians who can share the load, you might want to consider keeping your office open some evenings (until 7 p.m., say) and Saturday mornings. In addition to increasing patient satisfaction and loyalty, this can reduce patients' use of more expensive sources of care such as urgent care centers and hospital emergency departments. Moreover, it will turn out to be important for population-based care in later steps. As an alternative, or perhaps in addition, you might find it possible to set up a retail clinic linked to your practice or link to an existing clinic; that could be a reasonably effective after-hours extension of the office. As long as you have the necessary staff and physician availability, you could expand hours at any time. Still, it seems best to wait until the practice has successfully implemented the first four steps. The improvements they produce should be obvious to all and should motivate others in your practice to consider expanding hours. I'm not suggesting that individual physicians increase their work hours; this can be managed by adjusting start times to push some physicians' hours later in the day. I do think that primary care offices need to become available to their own patients at times that are less disruptive to those with school-age children and work responsibilities.
Step 6: Buy and implement an EHR
About six months after increasing the number of patients physicians see per day, you should have the cash flow to take on this expensive step. The hardware and software can cost up to $25,000 per physician, the maintenance is not cheap, and the new system will cut productivity until everyone is fully trained and used to it. However, because you have already created a team in which all members are working to the top of their licenses, the EHR will initially make itself felt mostly as a change in the way patient data is stored and retrieved. It is important to have nurses doing most of the work that doesn't require a medical license or skill in physical examination and diagnosis, and that includes entering data in the EHR system. In selecting an EHR, keep in mind that later steps involve using registries and creating electronic linkages with labs and other providers. You need an EHR that will be up to the demands you'll place on it down the road.
My rationale for delaying the purchase of an EHR until this far into the process is not just that you probably need to improve practice finances to make it affordable, but that the EHR will tend to reflect the way the practice operates. Building the team and its full functionality should precede the purchase and use of an EHR. The EHR system that seems adequate to you at Step 1 may seem completely inappropriate once you have a highly functional team. Moreover, depending on the EHR, the way you set it up may tend to codify and perpetuate practice operations, making it harder to build toward the highly functional practice you'll need if the implementation occurs too soon.
|Distribution of established patient codes||Net receipts||Improvement over ordinary practice|
|Ordinary practice; 20 visits per day (4,500 visits per year)||2%||15%||66%||16%||1%||$265,262||—|
|Improved coding (4,500 visits per year)||2%||10%||40%||43%||5%||$316,991||$51,729|
|Optimal coding (4,500 visits per year)||2%||8%||25%||60%||5%||$341,510||$76,248|
|Optimal coding; 25 visits per day (5,625 visits per year)||2%||8%||25%||60%||5%||$426,888||$161,626|
|Optimal coding; 30 visits per day (6,750 visits per year)||2%||8%||25%||60%||5%||$512,266||$247,004|
Step 7: Start doing systematic, population-based care
Once you have a high-functioning team and you've used your EHR system long enough to be comfortable with it, you're ready to start providing a level of care that few family medicine practices can attain today – building chronic disease patient registries, implementing some version of the chronic care model and developing proactive, population-level strategies for preventive care. Practices at this level of function are able to produce disease management and preventive care results far better than those of conventional practices.
Step 8: Buy and implement a patient portal
A portal can give patients new ways of interacting with your practice, help automate some tasks that may save staff time and even enable new billable services such as e-visits. A typical start-up cost is about $3,000, and monthly maintenance fees are about $100 per physician per month for a typical practice.
Portals are offered by some EHR vendors and by several companies that specialize in patient portals.13 The better companies do a good job of training and support. Depending on your system, this can allow your patients to enter their medical history in advance of the visit, do secure e-visits, access their lab results, access their personal medical record, schedule appointments and pay their balances online.
Step 9: Work with local health systems to create electronic linkages
If your EHR doesn't already include e-prescribing, interfaces with lab systems that allow automatic import of lab results into your EHR, and linkages to import documentation of care provided at other locations, now is the time to work on building these interfaces. They can be expensive enough to develop that it makes sense to put this step far along in the process, but they can greatly improve your capacity for care coordination and reduce unnecessary testing and treatments, not to mention reduce unnecessary data entry. Automating this can take enough burden from your support staff to add incrementally to your practice's capacity.
Step 10: Improve management of high-cost patients
Once you have working EHR interfaces with area labs and other providers, you are well situated to take another major step in delivering proactive care for patient populations. Partner with one or more local payers to identify the 5 percent to 10 percent of your patients who are generating 50 percent or more of your overall costs of care, and create strategies to help better manage care for this subpopulation. You can thereby reduce overall costs of care while improving the health of your sickest patients. Although you could probably do this earlier in the process, you will have the most complete and useful database to work from if you have the full integration with other care providers described in Step 9. Because this may require care management resources beyond your current staffing capacity, you may not want to take this step without receiving an additional care management fee or having some arrangement whereby you can participate in the cost savings you produce. It may also be more sensible to do this work as part of a larger consortium of practices, such as the highly successful Community Care of North Carolina (http://www.communitycarenc.com) that has greatly reduced state expenditures for Medicaid.
The following FPM articles, as well as the Anderson and Sinsky articles in the reference list, can help you implement many of the changes mentioned here:
“Coding Level-IV Visits Without Fear.” Waller TA. February 2006:34–38; https://www.aafp.org/fpm/20060200/34codi.html.
“Creating a Successful After-Hours Clinic.” Quackenbush J, Shenkel R, Schatzel V. January 2004:39–42; https://www.aafp.org/fpm/20040100/39crea.html.
“How to Select an Electronic Health Record System.” Adler KG. February 2005:55–62; https://www.aafp.org/fpm/20050200/55howt.html.
“How to Successfully Navigate Your EHR Implementation.” Adler KG. February 2007:33–39; https://www.aafp.org/fpm/20070200/33howt.html.
“Improving Chronic Illness Care: Lessons Learned in a Private Practice.” Mohler PJ, Mohler NB. November/December 2005:50–56; https://www.aafp.org/fpm/20051100/50impr.html.
“Making a Case for Online Physician-Patient Communication.” Adler KG. May 2008:A3-A6; https://www.aafp.org/fpm/2008/0500/pa3.html.
“Same-Day Appointments: Exploding the Access Paradigm.” Murray M, Tantau C. September 2000:45–50; https://www.aafp.org/fpm/20000900/45same.html.
“Working Together: Communities of Practice in Family Medicine.” Endsley S, Kirkegaard M, Linares A. January 2005:28–32; https://www.aafp.org/fpm/20050100/28work.html.
Essentials for the journey
There you have it: 10 steps to the kind of practice you might think you can only dream of. Needless to say, it looks a lot easier laid out in the compass of a short article than it will be to achieve in fact. But I believe this stepwise progression is not just possible but practicable. Depending on how and whether health care reform unrolls, I believe there are four keys to success:
First, family physicians must overcome the obstacles to coding and documentation that keep their current incomes far below what they might be. Since this problem has been apparent and commented upon for years, we clearly haven't effectively dealt with the obstacles to better documentation and coding, particularly the fear of a Medicare audit. It will be important for many practices to share their experience with appropriately documenting and coding at higher levels than are currently typical and successfully surviving an audit. My own recent experience and those I hear described by colleagues in Virginia who have passed audit tell me this is entirely feasible.
Second, family physicians will need to become expert at creating and sustaining a practice culture in which everyone feels safe, respected and valued. As the TransforMed program has demonstrated, practices who “get relationships right” (doctor-patient, doctor-staff and practice-community) are also the ones who tend to achieve the most elements of the PCMH.14–16 It may well be that the physicians who are drawn to family medicine are inherently ready and willing to do this but have become distracted, if not disheartened, by the growing burdens and diminishing payments that many now experience. If they are able to complete the first two steps of the progression, I believe they will recapture the joy of practice and, with it, the ability to be fully present again to their patients, staff, colleagues and families.
Third, payers will need to learn to welcome this sort of development in primary care. Currently about five percent of the health care dollar goes to primary care; if large numbers of primary care practices achieve high levels of proactive care, more of that health care dollar may go to primary care, and payers are likely to worry that this increase will result not in overall cost savings but in cost increases. However, there is good reason to believe that reductions in after-hours emergency department use, uncoordinated care with unnecessary testing and avoidable hospitalizations will make this model at least cost neutral.
It will be important to identify high-functioning practices and compare their performance with respect to overall costs of care, quality and patient experience. Innovative models described by Christine and Tom Sinsky and by Peter Anderson have clearly shown a greatly increased capacity and lead their peers when it comes to quality and patient experience of care.8,9 A collaborative effort by payers and health services researchers that demonstrated the expected overall costs savings, as state-level studies of Medicare beneficiaries suggest,17 would be very powerful in making the argument for rapid dissemination of this model.
Fourth, it will become increasingly important to create and sustain “communities of practice”18 that allow physicians and office staff from multiple practices to freely share the practical challenges they face and jointly create and test solutions, then disseminate those that work. In many ways, the “uncoached” arm of the TransforMed national demonstration project was such a community, in that they decided to share intimate details of their efforts to achieve idealized models of practice, sought and gave advice, and even came together in the same place to invest in creating and sustaining relationship with one another.19
While having a national clearinghouse of information and support such as the TransforMed program will be an enormous help to practices, I believe the creation of working collaboratives at the state and regional level will have the biggest impact on disseminating and sustaining models that work. State chapters of the AAFP can serve a key role in this regard, and I feel fortunate to have a visionary board in the Virginia Academy of Family Physicians (VAFP) that has helped the twice-yearly scientific assemblies of the VAFP become places where colleagues can share their successes and freely teach one another. Peter Anderson himself was one of these individuals three years ago, and his team care approach now has national attention. Academic units of family medicine with sufficient research infrastructure can help these collaboratives by providing health services research expertise when called upon, and consortia of highly functional PCMHs are ideal sites for teaching. Vibrant family medicine teaching practices in community settings have always been the best means to recruit students to our specialty.
It seems to me that we have a clear opportunity to make our practices and our health care system much better than they are now, and we may actually already have the needed resources. It's time to get started.