Stephanie Kim

Stephanie Kim, DMS I

Quality Improvement in Chronic Disease Management

Advisor: Nancy Morden, MD, MPH, Assistant Professor of Community & Family Medicine, Dartmouth Medical School

Summer 2009, DH-Family Medicine, Lebanon NH

Background

The issue of chronic disease management has become increasingly interesting to me as I move through medical school. My learning objective this summer was to learn about healthcare systems that promote the delivery of chronic care. I hoped to then use what I learned to understand how one could improve a Schweitzer project on diabetes management at Good Neighbor Health Clinic and family practices such as Family Medicine Associates in Springfield.

I first turned to the literature to understand theoretical models of chronic care. E. H. Wagner and others have described a chronic care model (CCM) with six components. “Self-management support” refers to formal programs that promote the patient’s self-management of their disease. “Decision support” refers to clinicians and other involved healthcare professionals, who are themselves supported by guidelines and medical education. “Delivery system design” is the proactive approach in managing patients with chronic disease. “Clinical information systems” include the patient registry, ideally electronic, which allows clinicians to view patient information in useful panels. A group of clinicians devoted to quality improvement make up “health care organization,” and “community resources” offer additional patient support outside the clinic. Though studies have shown that implementing some or all of these components effectively improves the outcomes of chronic disease patients, Wagner et al acknowledge that the CCM is not practical for exact replication from clinic to clinic. Ideally, general ideas taken from these six components would be turned into specific applications that fit the context of a particular clinic.

Still, in practice, it is difficult to build a system that treats chronic care patients, primarily because clinics have largely evolved to handle acute care. I was fortunate this summer to be involved with three clinics, which provided me with a spectrum of local healthcare systems. Good Neighbor is a free clinic that uses paper records, and, like most free clinics, is primarily set up to handle acute visits. Family Medicine Associates is a family practice in Springfield, VT that stores about half of its records on paper and half in the electronic database, DrNotes. The DHMC Family Medicine department at the Community Health Center uses the electronic record, Clinical Information Systems.

My summer project was a chart review of hypertensive patients at the Community Health Center for the hypertension quality improvement group. The hypertension group was a newly formed initiative that intended to make improvements in patient outcomes in the same manner as a recent diabetes initiative. They had found that using registries, flowsheets, and templates specific for diabetes visits had improved the tracking of patient progress, average patient HbA1c’s, and the consistency of follow-up visits and lab work. However, the needs of the hypertension population at the Community Health Center were less defined. A preliminary study characterizing the patient population was performed, finding that 2083 patients had two diagnoses of hypertension, with 42% of these patients having a blood pressure not at goal (140/90). Other preliminary improvements were made, such as moving blood pressure equipment in a more convenient location for the patient, and compiling patient education materials. However, the clinic did not yet have data that demonstrated the areas in greatest need for improvement. For the clinic, I compiled details from patient charts and provider notes in order to highlight some of the strengths and weaknesses of the current status of hypertensive care. I was able to explore the challenges of meeting the needs of chronic care patients, both inside the exam room and between visits.

Chart Review Summary

The goal of the chart review was to detect any gaps in the treatment of hypertensive patients, using standards from the JNC-7. Specifically, we wanted to assess how frequently blood pressure was being addressed at patient visits, as well as to assess the consistency in following up with care plans.

I compiled information from 56 patient charts. They each had two diagnoses of hypertension and a last blood pressure reading higher than 150 systolic and/or 90 diastolic. Every patient was last seen by a provider at Family Medicine between six and twelve months ago. Five patients from each provider were chosen randomly, unless the provider had less than five patients who met the criteria. The idea was to focus on a small group of out-of-control patients whom the clinic had not entirely lost contact with, but had, for some reason, difficulty being seen in recent history.

Worth mentioning is that the chart review was inherently biased towards more difficult patients. There were disproportionate hypertension panels for providers with very small or very large patient populations. For example, a provider with the majority of his or her patients in excellent control may only have had five patients that met the criteria of the chart review because they were particularly difficult to manage. The chart review thus is not meant as a reflection on the provider’s success at controlling his or her hypertensive patients. In addition, I also acknowledge that there is a discrepancy between what is recorded in office notes and what is actually discussed at patient visits.

I organized an excel worksheet by provider and compiled the last three care plans, taken from the clinical note, for every patient. I also listed the last three blood pressure readings in order to confirm that their high blood pressure was indeed a trend or to demonstrate that it was not. I listed other relevant information including age, gender, dates of the recent lipid panels, other active problems, active medications, and tobacco history. In a presentation to providers and staff at the Community Health Center during a weekly meeting, I summarized the information I had compiled into a powerpoint presentation. From the clinical note and appointment history, I noted whether a follow-up appointment was indicated, if the patient scheduled his or her appointment, and if he or she showed up. I also noted whether the patient was counseled on lifestyle changes, had a home cuff, what kind of visit the patient was seen for (i.e. acute visits vs. hypertension follow-ups or physical exams), and whether blood pressure was discussed during the visit.

The results suggested that the blood pressure and lifestyle issues were not being addressed as often as necessary. Family practitioners are extraordinarily busy in the number of issues they must discuss with a patient, particularly if they are in the schedule for a shorter, acute visit. However, a problem arises if the patient ignores his or her follow-ups for hypertension, because they are asymptomatic, and only comes to see his or her provider sporadically for acute visits. Acute visits may in many instances be the only chance in a while that a provider has to bring up a patient’s blood pressure. The results also suggested that, while a provider may have delineated a specific care plan and follow-up timeframe for the patient, the patient often does not schedule or show up for the appointment. Providers in many instances do not specify a suggested timeframe for follow-up, which substantially decreases the likelihood that a patient will return for a visit in an appropriate amount of time.

The discussion among providers and nurses following the presentation generated suggestions for improvement. For example, nurses should flag patients with hypertension on their charts so that the provider is reminded to bring it up during the visit. Providers should provide clear follow-up timeframes and/or consider providing a written care plan for patients to take home. It was suggested that home blood pressure readings were not being consistently entered into flowsheets. I hope that the information I presented served as a launching pad for further improvements in the management of hypertensive patients.

Further work

I learned many valuable lessons from compiling information for the chart review. First, I realized how important an electronic registry was to chronic care. It is vital to be able to pull data from a single patient or an entire registry to track the progress of a certain patient or the entire group of patients. Paper charts are simply much to unwieldy to review recent lab work, blood pressures, and appointment history in a quick, useful manner. Chronic care depends on the long-term tracking of patient progress, which realistically can only be managed through an electronic record. The importance of patient education also made an impression on me. A patient’s ability to self-manage their disease depends on and is greatly enhanced by how much they understand their illness.

With this in mind, I turned to Family Medicine Associates to examine how they managed their chronic care patients, especially since only half of their records were electronic. FMA received a grant to participate in a chronic disease initiative across the state of Vermont. Maureen Shattuck, the diabetes educator, showed me that the clinic uses a special electronic registry for their diabetic patients, the Vermont Health Record. This web-based registry is free for any clinic in Vermont, and provides users with visit flowsheets, graphs that track HbA1c and cholesterol levels, and means of detecting patients who have not been seen for awhile. Like the Community Health Center, FMA had done much work on diabetes management, but had not made as much headway with their much larger hypertensive population. Vermont Health Records has special panels for hypertensive patient, but FMA is far from entering that patient data. In the meantime, I drew up two posters on high blood pressure for the FMA, to promote their educational efforts.

What I learned this summer will likely have the biggest implications on the diabetes clinic that I manage with my classmate, Begem Lee. I realized that we cannot realistically move forward without a better electronic registry. We are now currently in the process of obtaining the Vermont Health Record system for the diabetic and hypertensive patients in the clinic. I have high hopes that an improved clinical information system will help us run the clinic more smoothly, track patients who may have fallen out of touch, and make it much easier for us to review who is due for lab work, eye exams, and regular checkups. I am also hoping that it will be a great tool for quality improvement measures. We already have patient education measures in place, but perhaps we could also improve access to other community resources.

Conclusions on chronic disease and primary care

This summer I hoped to learn about the systems of healthcare set up to treat chronic disease. What I actually learned is that, while there is much literature that describe theoretical chronic care systems and studies that have shown improved patient outcomes, it is an incredibly difficult task to modify practices in order to incorporate any aspect of the CCM. Perhaps the most important lesson I learned is that change is slow, but not because of resistant providers, difficult patients, or unorganized information systems. I found the Community Health Center to be an extremely well-run practice that is always trying to improve itself. Even at such a forward-thinking practice, however, it simply takes time to understand what exactly needs to be improved, what steps must be taken, and how improvements will be measured. It takes even more time for everyone to agree to and adopt new strategies. However, I am optimistic that real improvements will continue, and that we will eventually be able to curb the morbidity and mortality from chronic disease in our communities.

 FMIG Trip to Kansas City