EMRs and Primary Care: The good, the bad, and the challenges

One of the centerpieces of health reform as promulgated by almost everyone, and very much the Affordable Care Act (ACA) is the use of electronic medical records (EMR, also called, in a more inclusive formulation, electronic health records, or EHR). The Health Information Technology for Economic and Clinical Health Act (HITECH) specifically addresses specifications for EMRs. Demonstration of effective use of EMRs, including “e-prescribing” (in which prescriptions are routed electronically directly from the physician’s office to the patient’s pharmacy of choice), maintenance of patient registries (who in your practice has diabetes?) and compliance with a set of quality measures (What percent of the people in your practice with diabetes have had their sugar measured? What percent are in control?) account for a great deal of the added payment for chronic disease management, as well as payment for patient-centered medical homes (PCMH).

EMRs are a good thing for many reasons. At the simplest level, the fact that the records are on-line, rather than in paper charts, means that they don’t get “lost” and any doctor can see the notes of any other doctor. A number of years ago, prior to going to a real EMR, a large public hospital with many clinics where temporarily lost charts often meant that patient notes generated in one clinic visit were unavailable to another clinic, scanned literally millions of pages into a very basic EMR. While having none of the advantages described below, even this primitive method was a real step forward for them in being able to access the records. At their best, EMRs allow effective communication between doctors in a practice. For large multispecialty practices, this can also be between different specialists, and can even be integrated with the hospital’s medical record so that information from hospitalizations is immediately available in the same “chart”. The more that information is put in “digitally retrievable” format rather than free text, the more easily and thoroughly that a patient’s health trajectory can be understood. This is not only for numeric values, such as lab results and blood pressures that can be displayed on a flowsheet or graph, but even for history and physical items: Was that heart murmur present at the last visit? What is the history of the different medications that the patient has been on? Patient registries, as noted in the first paragraph, become an effective way of evaluating and improving the care given in the entire practice, not just for one patient, and are almost impossible without an EMR.

EMRs are not problem-free, however. The most common issue for physicians is that charting takes longer; filling in all this data takes time. This is worst when a new EMR is implemented, as old data has to be input (and this can even be when changing EMRs, not just going from paper, since of course they rarely “talk to” each other), but continues to be, on average, more time consuming than paper charting. In part, this may be because the notes are “more thorough”, or, looking at it the other way, that paper chart notes were inadequate. But it is also because the very structured nature of the EMR requires that a significant number of things be  entered/clicked (even to indicate “not applicable” or its digital equivalent) that would have appropriately not been mentioned in a paper note. Much of this added documentation goes beyond the information necessary to provide medical care for the patient, but is required to comply with government regulations and ensure that the document is “legally” sound. (It is a time long since the medical record existed solely as a reminder to the physician of what s/he had done for the patient in the past!) In addition, some of those regulations require the physician, as opposed to another health professional such as a nurse, to personally document certain items in the record, often to a degree that seems unreasonable to physicians.

There is an ironic turn to this. Most discussion in public policy circles is directed to increased inter-professional function and team work, as characterized by the patient-centered medical home. In part, this is because the current and projected shortage of primary care physicians means that there is no way that they, working alone, will be able to meet the health needs of the American people; if they are already working on a “hamster wheel” (see Family Medicine in the Era of Health Reform - 3, May 23, 2011), the changes described by Phillips (see Primary Care, Medical School Debt, and US Health Needs: Analysis from the Graham Center, May 30, 2011 ) and discussed in detail by Margolius and Bodenheimer[1], will increase the burden beyond any hope of sustainability. In addition, an effectively functioning team of health professionals (including nurses, pharmacists, social workers, and others[2]) makes for higher quality care. This is very clearly articulated in Dr. Atul Gawande’s recent address to the Harvard Medical School commencement, “Cowboys and Pit Crews” published on his New Yorker  blog. The irony is the increased requirements mentioned above, sometimes explicitly stated in law, but often in federal regulations and most commonly by Medicare “carriers” and interpreted by institutional compliance officers, have increased what the physician, him or herself, needs to document in the medical record (and, by implication, have actually done him or herself). These requirements both decrease effective team function, and increase the burden of electronic charting.

Thus, the ability of the EMR to record, and hopefully make retrievable, large amounts of data, raises the expectation that that data will be inputted, and also allows monitoring to ensure -- in the cases where lawyers or compliance officers have concern – which individual is doing it. In any new technology that increases the ease of accomplishing something, or the availability of a person or data, there is the corresponding tendency to expect it; this often has the ironic effect of increasing, rather than decreasing, workload. The internet and email allow us to work from home; cell phones, pagers, and email all increase our availability even when not at work or at home. This allows us more flexibility, but it has also led to the expectation of immediate access and, for many professionals including physicians, the virtual elimination of the concept of “work” versus “off” hours. The electronic medical record allows me to chart from home – or anywhere I can get an internet connection – and so I do.

The introduction of a new EMR, already a complex, difficult and daunting process, is also often used to change workflow, the processes by which the work of the practice is accomplished. This is virtually always a mistake, for providers and staff must now learn not only how to navigate and document in a new and strange systems, but to do things in a completely different way. The changes may be desirable, or they may not be, but certainly will require time and effort to identify whether they are, work it out, and “get it right”. Thus, it is greatly preferable to change that workflow prior to institution of the EMR (or if necessary after it is successfully adopted); the alternative is that learning new processes, especially when they are poorly conceived, gets lumped in with and blamed upon the EMR, increasing resistance to its adoption and potential benefits.

An interesting, and perhaps important, sidelight of the introduction of the EMR in our family medicine clinic was that the implementation team, composed of experts from the computer company and “superusers” of nurses from our group practice, saw how much more complicated the practice – and thus the documentation – is in primary care than in other specialties. In most sub-specialty practice, a few diagnoses -- and thus a few types of workflow and documentation strategies -- account for almost all visits, while in primary care the breadth of encounters (acute/chronic, prevention, adult/child/pregnancy, medical/psychosocial) in a single session, combined with the complexity of dealing with multiple chronic conditions based in a variety of organ systems rather than one, is actually breathtaking (see, for example, Primary Care: What takes so much time? And how are we paying for it?, May 21, 2010, "Uncomplicated" Primary Care?, Oct 8, 2009). Contrary to what they had been led to believe, they discovered that primary care was harder and more complex and more difficult to document – and of course required seeing more patients in shorter amounts of time for less reimbursement (which also leads to an ability to afford fewer support staff). This team, at least, gained a new respect for what primary care practice involves.

As we inevitably and inexorably move to reliance upon EMRs, we must be on guard to resist all the temptations to load every possible expectation upon them, and upon the providers who use them. They have enormous potential to not only increase quality but to increase teamwork and communication, and to even be labor-saving, but only if used wisely and judiciously.

---

[1] Margolius D, Bodenheimer T, “Transforming primary care: from past practice to the practice of the future”, Health Aff (Millwood). 2010 May;29(5):779-84.

[2] Even, for example, lawyers. A number of practices, particularly in academic medical centers (including our family medicine clinic at KUMC) have legal partnerships where lawyers (often from Legal Aid) and law students help patients with legal problems they could not otherwise get help with, right in the clinic. It is amazing how often a person’s health improves when they no longer are as worried about their immigration status, getting evicted, receiving benefits, or the implications of divorce, among other issues.

Reinventing Primary Care: Themes and Challenges

.
My May 27 entry, Universal Coverage and Primary Care: The US needs both, focused on one article from the May 2010 issue of the journal Health Affairs. That article, by Barbara Starfield, (“Reinventing Primary Care: Lessons from Canada for the United States”) was just one of 65 articles in this issue that addressed the topic of “reinventing primary care”. The articles cover the gamut of the history of primary care, the problem of undersupply of primary care providers, proposals for increasing the number of primary care providers, analyses (such as Dr. Starfield’s) of the health systems of other nations’ that are based on primary care, proposals for how the organization and structure of primary care practices need to be changed, and case studies of models of practices and demonstration projects that have implicated innovative approaches to primary care delivery. Many of these practice changes are contained under the rubric of the Primary Care Medical Home (PCMH), which is also the abbreviation for the related Patient Centered Medical Home. Indeed, the Patient Centered Primary Care Collaborative (PCPCC), a coalition of major employers, insurers, providers, pharmaceutical and device makers, and consumers, uses both “PC”s in its name (one hesitates to say that it is thus very PC!)

One of the articles is in fact “The multistakeholder movement for primary care renewal and reform”, by Paul Grundy, et. al., specifically addresses and discusses this collaborative, as well as its proposals for change in the entire structure of the primary care delivery system. Grundy, who is vice-president for global health reform of IBM, is president of the collaborative. It has brought together these various stakeholders in recognition not only of the overwhelming data that shows systems that are built upon primary care are more efficient, more cost-effective, and lead to better health outcomes, but on the actual experience of IBM and other multinational companies. These companies find that their health costs, and the health (and thus lack of time off for illness) of their workers is dramatically lower in countries in which the health system is built upon a primary care base. It is actively involved in educating, advocating, and demonstrating the importance of developing such a base in the US.

Many of the other articles in the journal address changes that need to happen to allow the small (and, as I have pointed out, likely to stay too small even if there is a significant increase in production from our current paltry 16%) number of primary care providers to care for larger panels of patients, while maintaining or increasing quality of care, patient satisfaction, and efficiency. Two of these are co-written by Thomas Bodenheimer, MD, ‘Primary Care: Current Problems And Proposed Solutions”, by Bodenheimer and H.H. Pham, and “Transforming Primary Care: From Past Practice To The Practice Of The Future”, by D. Margolius and Bodenheimer. Another perspective, more from that of the individual physician than the overall health system, is Lawrence P. Casalino’s contribution, “A Martian’s Prescription For Primary Care: Overhaul The Physician’s Workday”

What is striking about these articles in the similarity of their recommendations. The recurrent themes include the need for multi-disciplinary teams of health professionals who all play roles in caring for patients, and panels of patients. This goes beyond the simple “nurses doing callbacks” to patients before, or instead of, the physician. It means that nurses provide the care that they can, that pharmacists and psychologists and social workers all are part of the team, communicating with each other but often operating independently. Group visits are another theme; often people with the same (or, in fact, different) conditions can benefit from being seen in a group. This can be for a more formal didactic session of patient education about their condition by a nurse, or pharmacist, or health educator, or physical therapist, or physician, or medications, or other treatments, often combined with a great deal of person-to-person interaction. There is more to this than efficiency; people actually benefit from the fact of being in a group, of sharing experiences, and ideas, and successes and failures.

Another important and recurrent theme is that of physician-patient interactions that do not involve face-to-face contact, but rather phone calls or emails. Again, this is not just a matter of efficiency for the physician; patients often have concerns that can be successfully addressed by one of these other methods that do not involve them having to take off work, drive a long distance, look for parking, and wait in the waiting room. A physician can be far more effective, and interact with a much larger number of people, if an afternoon consists of seeing a few in person, a much larger number by telephone, and an even larger number by email; a number of products exist that provide not only secure email communication, but provide a structure for the patient to supply information that will help the doctor (or NP, or nurse, or whoever on the team is most appropriate) provide the greatest help.

In his article, Casalino lists five reasons why a physician should see someone in person:
“(1) for a first visit; (2) when it may be necessary to engage in some physical maneuver for diagnostic purposes—such as palpating the abdomen, listening to the heart, or performing a skin biopsy; (3) for specific therapeutic purposes, such as injecting a joint; (4) when the patient has problems for which lengthy discussion would be helpful; (5) when for psychological or emotional reasons it seems better to see the patient face-to-face; and (6) when face-to-face visits are necessary to build trust.”
These are very good, but I would simplify it even more: the physician should see the patient face-to-face when either the patient or the physician think it is important.

The effectiveness and satisfaction from increased phone calls substituting for visits was clear in the late 1990s when capitation, rather than fee-for-service, was a dominant mode of payment (of course, at that time the internet was not yet developed enough for most people to be using email). To reprise, and to improve upon that process will, obviously, require a reimbursement system that does not pay only for face-to-face physician visits. This is another common theme to many of the articles in this issue. It is also not happening in most places. It is, however, a sine qua non for such practice reorganization. And increasing the primary care supply.

Which of these themes is the most important: increasing the supply of primary care physicians, reorganizing practices to become true Patient Centered Medical homes, utilizing all of the strategies above and more, or restructuring the way health care is paid for? They are all important, all related, and all dependent upon each other for success. What is not addressed in most of the articles in this issue of Health Affairs, however, is arguably the most important: ensuring health coverage for everyone. The work of the Patient Centered Primary Care Collaborative may go a long way to having a more rational delivery system for those with access, but as long as the same old hands, in particular the insurance companies but also the drug and device makers, are part of the decision-making process, they are unlikely to come up with a plan that truly covers everyone, single-payer, “Medicare for All”, or any other rational system. And without that, all the primary care reform is not going to really work. It is not only a moral issue (although that should be enough!); it is that the cost, in terms of work hours lost, unnecessary suffering, direct dollars spent on diseases that have advanced too far because people have delayed care, and ultimately worse outcomes, is unsupportable and unsustainable.

We’ve passed PPACA. Now it’s time for real health reform.
.

Primary Care: What takes so much time? And how are we paying for it?

.
In a piece that has gotten a lot of attention, “What’s Keeping Us So Busy in Primary Care? A Snapshot from One Practice”(New England Journal of Medicine, Apr29,2010;362(17):1632-6), Philadelphia general internist Richard J. Baron writes about the many tasks – many of them unreimbursed – that occupy the time of the physicians in his practice. While the physicians in the practice saw an average of 18.1 patient visits per day (the one activity they were paid for), they also returned an average of 23.7 telephone calls, and 16.8 email messages. They refilled 12.1 prescriptions, reviewed 19.5 laboratory reports, 11.1 imaging reports and 13.9 consultation reports per day. They also filled out large amounts of paperwork that they do not report on because they are not captured by their electronic medical record, such as “…administrative forms (e.g., for physical examinations for work, camp, and school and Family Medical Leave Act forms), correspondence received from health plan (e.g., disease-management letters), and reports on home care and physical therapy.” That’s a lot of work that is not being paid for. This is a huge issue in providing primary care, and one that is not often appreciated by the subspecialists, medical and surgical, who are very highly reimbursed for the procedures that they do, so that the paperwork, phone calls, etc., such as they are, are well subsidized.

Dr. Baron’s report reinforces both frequent observations and studies that have been done earlier. In 2003, Yarnall, et. al., published an article that looked at the amount of time it would take a primary care physician with a typical practice of 2,500 patient to provide all of the preventive care recommended by the US Preventive Services Task Force (USPSTF) at the “A” or “B” level. In “Primary Care: Is There Enough Time for Prevention?” (Am J Public Health. 2003;93:635–641), they found that:

“In all, an annual total of 1773 hours, or 7.4 hours of every working day, is required for the provision of all recommended preventive services to a practice of 2500 patients with age and sex distributions based on the US population.”

That is a staggering statistic; in itself this is a full-time occupation, and yet it does not include any time for managing the chronic diseases that people have, or the acute conditions that are bothering them! In a later piece in 2009, “Family Physicians as Team Leaders: ‘Time’ to Share the Care” (Prev Chronic Disease Apr2009;6(2):A59), Yarnall and colleagues further assess all three types of care (preventive, chronic, and acute) and determine that together they would take 21.7 hours a day! This is reassuringly less than 24, as even doctors need to sleep and eat!

The recommendations from Yarnall, et. al., and from other recent pieces assessing the changes that will need to take place in the structure of primary care, coincide with the decisions made by Dr. Baron’s group to increase the size of their team. While the internists in Philadelphia mainly added a triage nurse, a true team requires a more comprehensive approach. Proposals planning for the Patient-Centered Medical Home (PCMH) understand that “it takes a team,” not just a physician. Among the best recent review and analyses is “Transforming Primary Care: From Past Practice To The Practice Of The Future” by David Margolius and Thomas Bodenheimer (Health Affairs May 2010, 29(5): 779–784 ). Virtually all these proposals advocate developing a team of providers, including doctors, nurses, medical assistants, pharmacists, secretaries/clerks, social workers, mental health specialists, even public health professionals, to collaboratively provide the appropriate and necessary care.

This is a great idea, but hard to implement in the current fee-for-service system, a system unlikely to change given the degree to which the new health reform law is built around insurance companies. Dr. Baron’s practice was participating in a demonstration project to provide coordination in the interest of the PCMH that was funded by the state of Pennsylvania; most practices are not. As I have said before, if proceduralist could do the procedures and subspecialists manage their particular rare or advanced disease, and allow primary care doctors to manage complexity and counsel and do prevention (and mental health professionals, and social workers, and pharmacists, and therapists to all do the things they were best able to do) rather than trying to do more procedures because they would make more money, this would make – sense.

Comprehensive care for people is a big undertaking. It is much more than consulting on one problem, or doing a single procedure and follow up. It is caring for many problems, and providing the preventive care that people need, and addressing their acute complaints, and doing counseling for both psychosocial issues and decision-making issues (e.g., “should I have this surgery?” “what is the risk of this diagnostic procedure recommended by the consultant?”) It is being available to refill lost or expired prescriptions, reviewing lab and imaging results and consultant reports, and answer questions by phone or email, and review and coordinate the care being provided by consultants. It is doing all those things that people expect that their “family doctors” (in whatever specialty they are actually certified) will do. It is a very different practice than that of a subspecialist; while for the latter “the buck stops here” for treatment decisions for a particular condition, for the former “the buck stops here”, period. There is nothing in the medical realm with which they cannot, reasonably, be expected to be involved.

Consider a patient who tells a subspecialist, say their cardiologist, about their knee pain. The cardiologist says “I don’t do knees,” and sends them back to their primary care doctor, or refers them to an orthopedist. On their next visit the patient says “I saw the orthopedist you sent me to,” and the cardiologist, reasonably, says “I don’t do knees. Do whatever s/he told you.”

Now consider the patient presenting with the same knee pain to their primary care doctor. Maybe that doctor has a good idea of what the problem is and how to treat it, but to be sure refers the patient to the orthopedist. On the return visit, the patient now wants to talk with “their” doctor about what the orthopedist said, what s/he recommended, what the primary care doctor thinks about that, and needs help making a decision. So, while for the subspecialist (cardiologist, in this example) a referral disposes of potential work, for the primary care physician, it will usually, while reassuring the doctor that the right care is being done, actually increase the amount of work.

Most primary care practices discover, as did Dr. Baron’s, that it is inefficient to have physicians doing virtually any of the work that could be done by a lower-paid staff member; this could be a nurse (which his practice did not employ initially) or a medical assistant. Subspecialists have always known this, employing nurses, physician’s assistants, and others (including medical “fellows”) to augment their productivity; for example, a surgeon can be in the operating room while a physician’s assistant sees patients in the office. But it takes money to pay such staff.

As long as the only reimbursement is for actual patient visits, and that reimbursement is spectacularly lower for primary care doctors than for subspecialists, and particularly proceduralists, the financial viability of such practices will be low, and the attractiveness to highly indebted medical students comparably low. This calls for a restructuring of the entire payment system, into one that encourages collaboration and the most appropriate management. A capitated payment system is very desirable because then patients can be “seen” in the most appropriate way based on the perception of patient and physician – phone, email, office, hospital, home – but only if that reimbursement is high enough to make the development and implementation of patient centered systems worthwhile.
.