Primary Care Advocacy Tool Kit

PATIENT VIGNETTES

Understanding Primary Care

Vignette 1: Access to Health Care
Vignette 2: Specialty Bias
Vignette 3: Continuity and Coordination of Care

Vignette 1: Access to Health Care

A 22 year old office secretary develops a severe sore throat and fever in the evening. She awakens in the middle of the night feeling worse and calls her physician’s 24 hour emergency line, as instructed by her HMO. The covering physician assures her that the problem can wait until morning, and that if she calls back between 9 and 10, it is almost certain that her physician or nurse practitioner could see her that day.

Newly employed and on a probationary period, the woman is unhappy with the prospect of missing a day or half day of work and pay. She considers her options:

1. Go to the emergency room now and pay the $25 co-pay.
2. Do what the covering physician suggests.
3. Take some Tylenol or aspirin and see if she gets better on her own.

Possible scenarios:

1. She goes to the emergency room. Not knowing the patient, the provider prescribes antibiotics to protect against the possibility that this is strep pharyngitis and that information from a culture would not reach her primary care provider.
2. She goes to her doctor, who does a culture and manages the sore throat based on the result.
3. She goes to work without getting care and is feeling fine within 5 days.

Negative potential consequences of each scenario:

1. The majority of such cases are not strep. The antibiotics are likely unnecessary. Prescribing them introduces the risk of serious allergic reactions, which can rarely be fatal. In addition, overuse of antibiotics increases the rate at which bacteria develop antibiotic resistance, a threat to public health. The visit costs the patient $25 instead of $10 and costs the insurer as much as three times what an office visit would cost. The visit may result in less patient education about the best management of this problem in the future.
2. She loses a day’s pay, and in a worst case scenario, gets cited for absenteeism and loses her job after her probationary period.
3. She is in the minority who have a strep throat, and, although she recovers from the sore throat within a few days, she develops rheumatic fever as a rare complication of untreated illness, and eventually becomes disabled due to heart disease.

This example illustrates the difficulty our health care system has in providing care that meets everyone’s needs. Patients and providers both want to have high quality care provided that protects patients from harm, but because patients do not always know the potential consequences of their illness, and because providers may not take into account the social and economic consequences of their care plans, there is a possibility that sub-optimal care will be given.

The structure of the care system can increase or decrease this likelihood. System components that increase the likelihood of one or more bad outcomes in this case include:

1. Inadequate communication systems between emergency and primary care.
2. Economic incentives designed to direct patients to primary care, but that do not take into account the full cost-benefit equation for patients.
3. Health financing that confuses costs with charges. Although emergency rooms are thought of as both less appropriate and more costly, in this case, since the emergency room is already open, the marginal cost of seeing the patient may be as little as the marginal cost of the office visit. This cost confusion precludes discussion about the possibility of providing appropriate care in this setting.
4. Primary care systems that do not provide for adequate off-hours care.
5. Real and perceived lack of accommodation by employers for health care needs of employees

Solutions to providing more optimal care in this case could include:

1. Robust information systems, so that a patient seen in the emergency room could be assured of having information transferred to a care provider for proper follow-up.
2. Development of more urgent and off-hours care for routine problems that could be offered in settings intermediate to emergency rooms and physician offices.
3. Financing systems that align charges with costs. This would likely require major reform, such as a single universal payment system.
4. Laws and regulations, or cultural change, that protect worker income and status when seeking care according to appropriate recommendations.

Getting to such solutions will require stakeholders in the health care system to jointly discuss and understand vignettes such as the one presented. The current system encourages sub-optimal care because the focus of health policy has been to create economic incentives to induce the patient and provider to accept the service with the lowest charge, without reference to the potential negative clinical and public health consequences of the actual behavior such incentives produce.

Vignette 2: Specialty Bias

A 13 year-old boy noticed a red, itchy, scaling rash that developed between his toes. He mentioned this to his mother. The boy's older sister suffered from moderate acne and therefore was seeing a dermatologist. His mother asked the dermatologist if he could see the boy on the same day as a visit by the sister. The dermatologist agreed. The dermatologist took a fungal culture of the rash, and when that came back as negative, he diagnosed eczema and prescribed Valisone Cream, a potent topical steroid.

The rash improved but never disappeared, and for the next 6 years would get better and worse with no clear pattern. The boy continued to use the cream, which usually had a soothing affect and reduced the intensity of the rash when used consistently, but the rash continued to spread over the entire foot over time and also appeared on the other foot, and at age 17 also appeared in the groin. The dermatologist continued to prescribe the steroid cream.

At age 19, the boy went on a 10 week backpacking trip through Europe, where (on his student budget) showers were few and far between. The rash became severe and covered much of his feet, thighs, and groin. On return, he consulted the student health service of his university, where the correct diagnosis of a fungal infection of the skin (which had started as athlete's foot) was made. The steroid cream was changed to a topical fungicide, which initially cleared up the rash in three weeks.

The infection by this time has spread into several toenails. In addition, over the years of steroid treatment the infection had penetrated deeply into the thickened skin of his feet, and so the itchy rash would reappear on the boy's (now man's) feet whenever he stopped the medicine for more than a couple of weeks. As a result, he required the use of hundreds of dollars of medication per year for 20 years. Only when safe and effective oral fungicides were developed and marketed, some 25 years after the rash first appeared, was he cured of what had started as a simple case of athlete's foot.

Analysis

Although it could be argued that the dermatologist made a significant error, this case also demonstrates the problem of specialty bias.

Specialty bias reflects the natural tendency of people to generalize, and to assume that what they normally and usually see will also be what they see even if they are faced with a different set of circumstances. Although athlete's foot is by far the most common cause of a foot rash in a 13 year old boy, specialists see it far less often because most teenage boys would either self treat the problem at home, or seek care from their primary care provider. A specialist is much more often confronted with the unusual – for example a rash that does not respond to usual treatment prescribed by the primary care provider.

In this case, the boy bypassed the primary care system, not because it had failed to make a proper diagnosis and treatment, but because of the circumstance of his sister already being in specialty care.

No test is perfect – and so it must be interpreted in light of the probabilities of disease that exist before the test is done. In the community, perhaps 99 out of 100 foot rashes in a teenage boy are athlete's foot. If the test if falsely negative 5 times out of 100, then a negative culture is roughly 5 times as likely to represent an incorrect test then a different diagnosis (for every 100 boys, you will get approximately one “true negative” and 5 “false negatives”). If a dermatologist, however, sees rashes that for the most part have been unsuccessfully managed by home care or primary care, then fewer of them will be fungal infections. If, for example, only 50 out of 100 such rashes were athlete's foot, then a negative test is more likely to be correct than not (roughly 20 times as many of the negative tests will correctly identify someone without athlete's foot as will be wrong).

The danger of specialty bias occurs when a specialist sees a patient who might normally be seen in a different setting – and the specialist fails to make the mental adjustment that they are seeing someone from the general population, who is much more likely to have common presentations of common problems, than from their more familiar specialty world, where rarer diseases, and rarer manifestations of common diseases, are the norm.

There are two ways to overcome specialty bias. One is proper and ongoing training of specialists to make sure they remember the potential for bias. This is part of proper training, and is frequently done, but unfortunately the natural human tendency to generalize is so powerful that such mistakes will happen and do happen even with the best training. Training will reduce but not eliminate the problem.

The second way to overcome specialty bias is to use the primary care system. Most illnesses that people develop are common and present in common ways. These are best evaluated first by a generalist, not simply because a generalist can make the proper diagnosis and treatment, but also because the generalist will be more likely to make the proper diagnosis and treatment because they are seeing the things that they see all the time!

Sometimes it's clear from the beginning that a specialist is needed. Our system could do better at finding ways to allow patients to bypass unnecessary visits. At other times, people do have conditions that generalists fail to diagnose properly, or that they do not have the expertise to treat. Generalists are trained to recognize their own limits, but again the system could do better – first to make sure that generalists have this training and second to make sure that there are no inappropriate incentives to limit timely referral.

Nevertheless, despite the public fears, fueled by managed care restrictions, that people are denied proper access to specialists, we must remember the converse is also true and may be a greater threat – people are being denied the opportunity for proper evaluation by a generalist who will be more likely to properly manage the more common problems in a correct and low risk fashion.

Vignette 3: Coordination of Care

A middle-aged man who was known to have coronary artery disease needed an elective surgery. He had never had a myocardial infarction (heart attack) but previously had stable angina pectoris. Nine years before his surgery he had a coronary arteriography which had shown mild to moderate disease and no evidence of critical stenosis (narrowing) of the arteries. His angina had subsided. Over the two years before his surgery, he had been having some chest pain which he had discussed with his primary care provider. His provider had reached the conclusion that his chest pain was not related to his coronary disease.

Because of the chest pain, the surgeon who was set to operate on the patient requested a cardiology consult. The cardiologist recommended a stress test to make sure there was no active coronary disease. The test was delayed, however, and the patient became anxious and developed chest pain. The stress test was cancelled, the patient transferred to a telemetry unit. Here the patient became increasing anxious, his chest pain worsened, and a electrocardiogram showed some equivocal changes. He was transferred to intensive care, treated aggressively for a possible infarction, and underwent arteriography.

The patient did not have an infarction, and his arteriogram showed mild improvement compared to the one from 9 years before. Fortunately, the patient did not have any serious adverse reactions to the medications given or to the arteriography, but his surgery had to be cancelled and rescheduled for several weeks later.

This case illustrates the importance of communication between patient and provider, and among providers.

The “expert” cardiologist had access to the best tests and technology, but did not have a long standing relationship with the patient, and therefore did not know what the patient’s original anginal pain was like and did not know whether the currently reported chest pain was different from that original pain. He/she did not have a clear sense of the patient – whether this was someone who was stoic, or someone who tended to report every minor ailment. He/she did likely did not know the quality of effort that led the primary care provider to conclude that this new chest pain was not heart disease.

A good cardiologist could partly overcome this lack of long term relationship and the detailed knowledge that comes with it through a careful history, but such histories take a long time to elicit, and it is common (if less than ideal) practice in rapid assessments such as a pre-operative evaluation to rely on testing and technology, rather than what is viewed as the softer science of the history. Tests can often be incorrect, however, and they are much more likely to be incorrectly interpreted if a careful history and judgment about the likelihood of finding disease is not addressed before the testing is done.

The best care in this case would involve communication between the primary care provider and the cardiologist. The information gained over time by the primary care provider can be reviewed and challenged by the specialists who become involved in the patients care, but that information should not be lost. In this case, that information might have saved the patient unnecessary tests, treatments, and risks.

Adapted from:

Mold, JW and Stein, HF, “The Cascade Effect in the Clinical Care of Patients” NEJM 314(8); Feb 20, 1986: pp 512 - 514