Medical Decision Making

Principle #1

Not all patients with a given diagnosis are of equal risk for complications:

Being aware that diseases fall on a spectrum of severity and are dynamic are important starting points in approaching treatment decisions. Q5.1 This means that clinicians cannot use a ‘one-size-fits-all’ treatment plan. The writers believe this often leads to inappropriate treatment. It is necessary to weigh the risk of disease versus the risk of treatment for each patient individually to determine who should be treated, recognizing that these variables may change over time. For example, not all patients with a diagnosis of lupus erythematosus require prednisone or hydroxychloroquine. Determine where the patient falls on the spectrum of the disease activity to guide this decision making. For instance, is the disease restricted to the skin? Does the increased photosensitivity limit the patient’s daily life activities? The answers to these types of questions will help formulate an appropriate plan. Clinicians should continue to reassess patients who may go through periods that require more intensive treatment and times that they do not.

Principle #2

Patients will vary in the amount of autonomy they wish to have in the decision-making process, but it is important to allow them the opportunity to be involved in decisions for their own care:

It is imperative that the patient be involved in the decision-making process because, ultimately, the patient is the one directly impacted by the decision and thus must understand and agree with the process. The amount of autonomy that individual patients wish to have will vary depending on the person and the decision being made. It is up to the physician to navigate where each patient falls on this spectrum. Many patients will be more passive, deferring to the physician to make all decisions based on their professional judgment. Conversely though, some patients will want to be more actively involved and lead the decision-making process. Regardless of where each patient falls on this decision-making spectrum, they must at minimum understand their disease process and risk involved in general. This is not only important for building trust and rapport, but also for increasing patient compliance. Patients must be provided with all realistic diagnostic and treatment options available before making a decision.

Principle #3

The spectrum of benefit must be expanded to include consideration of quality of life for the patient:

For many patients, quality of life while undergoing the desired intervention can be just as important as the efficacy of the treatment. This takes into account (1) the amount of monitoring required, (2) the route of drug administration, and (3) their likely tolerance of the treatment. All three aspects should be described in detail to the patient before initiation of treatment. Based on the authors’ experiences, often it is not just the AE that deter a patient from selecting a certain treatment but instead the impact it has on their quality of life. An example of this is the frequent initial laboratory monitoring needed when initiating dapsone therapy. The cost and time of having laboratory tests completed biweekly to monthly may prove challenging for some patients. Being clear with these expectations before starting therapy will ensure higher compliance as well as improve patient and physician satisfaction.

Principle #4

When selecting therapy, consideration should be given to patient comorbidities and personal preferences:

The prescriber should narrow therapeutic choices based on comorbidities and patient preferences. For example, the authors avoid methotrexate therapy in patients with liver disease, significant alcohol history or renal insufficiency. In addition, it is wise to avoid tumor necrosis factor (TNF)-α inhibitors in patients with a history of lymphoma or moderate-severe congestive heart failure. We also consider patient tolerability such as preference of route of administration, dosage frequency and vehicle of topical medication. Consideration to these details can result in significantly improved patient compliance and satisfaction. For most conditions, there is not a therapy that is ‘one-size-fits-all’ and patients must be paired with a regimen that maximizes effectiveness and safety.

Principle #5

There is power in being able to say to the patient ‘I do not know:

Trust is essential in the physician–patient relationship. If the patient does not trust the provider, the patient may not provide the history that would lead to an accurate diagnosis. Without trust, patients will likely not be compliant with the outlined treatment plan. This will result in less than optimal treatment outcomes and fault will ultimately be placed on the physician.

Ways to strengthen the physician–patient relationship include active listening, setting agendas during appointments, empathizing and placing an emphasis on patient education. Q5.2 However, there is also power in being able to admit what you do not know. Patients will generally appreciate your honesty, and it will help build a trusting relationship and allow you time to further research the topic before providing the patient with a ‘final’ answer. If the answer is not definable, providing reassurance that you do understand the treatment risks goes a long way in building trust. For example, if you have not been able to give a patient a specific diagnosis, provide reassurance that you definitely know the potential serious causes and that these have been excluded.

Principle #6

Use the medical decision-making ‘triad’ to arrive at a decision point for patient care:

• 1.

  Q5.3 Scientific literature : There are few topics in clinical medicine in which everyone is in 100% agreement so each provider must decide how much ‘certainty’ is enough. This will be further detailed later in this chapter, but a preponderance of evidence in one direction is the realistic goal.

• 2.

  Biologic plausibility: As adequate data is often lacking, we must rely on our own judgement to make a medical decision. Simply put, the mechanism of disease should correlate with the mechanism of treatment. Many times, these decisions can be made after clinical exam for the experienced physician with or without the assistance of a skin biopsy. For instance, diseases with a predominance of neutrophilic infiltrates such as pyoderma gangrenosum, subcorneal pustular dermatosis, and Bechet syndrome should respond to antineutrophilic agents such as colchicine or dapsone. Another example would be autoimmune diseases with lymphocytic infiltrate such as lupus erythematosus or dermatomyositis. These diseases typically respond to (1) antimalarials which lead to a corresponding decrease in the formation of major histocompatibility complex (MHC)-protein complexes that are required to stimulate CD4+ T cells, or (2) immune modulators such as methotrexate.

  Conversely, when a clinician is presented with an adverse event, the mechanism of the drug should correlate with the mechanism of the AE being experienced. For example, if a patient develops ichthyosis while on a statin medication, it is reasonable to suspect the statin as the cause because they lead to the depletion of cholesterol in the epidermis.

• 3.

  Personal experience: This can be either (1) direct, meaning one’s own experience, or (2) indirect, meaning a colleague’s or mentor’s experience. This can be invaluable, as much of medicine is an art more than science and can be improved upon over time. The more you see, treat and follow a certain disease, the more of an expert you will be become. A pitfall, however, occurs when a clinician identifies an area of expertise and develops ‘tunnel vision’. This can lead them to favoring what they have seen before, thus creating potential bias. Conversely, a provider may overlook a correlation because they have not seen that specific outcome before. For example, severe ketoconazole hepatocellular toxicity is a 1 in 10,000 risk, so a provider will need to see 30,000 cases treated with ketoconazole before one’s personal experience is ‘statistically significant.’ Personal experience is often not as definitive as we believe and should be used alongside scientific literature and biologic plausibility.

Principle #7

Correlation in timing does not establish causation:

Q5.4 When two events coincide with each other they ‘correlate’. but this does not mean that they are directly related; thus, this simultaneous timing does not prove ‘causation.’ This principle can be applied to MDM when clinicians are trying to determine both the cause of the efficacy of an intervention or when determining the cause of an adverse event. This chapter will concentrate on defining causation of benefit, but the principles may be applied to both scenarios.

The gold stand for determining causation of benefit is a double-blinded, randomized controlled trial. If this is available, great. But unfortunately, the majority of the time this is not available, so the clinician is left to carefully consider what literature is present. Evaluating medical literature is a critical tool in MDM.

If interventional studies show significant improvement, then this confirms A causes B.

For instance, a pharmaceutical company compares a new biologic medication versus placebo for the treatment of psoriasis. If the study shows marked improvement of psoriasis with the biologic and not the placebo, the study could suggest causation. When interpreting studies such as this, it is important to keep all other variables as uniform as possible. For instance, assess whether the study controls for additional therapies being utilized, disease severity, patient population, etc. This will provide more certainty when determining causation.

If interventional studies do not show correlation, they cannot prove causation. They would only show chance occurrence.

In this scenario, vitamin D ₃ ’s effect on psoriasis was studied. The study grouped patients by disease severity to add uniformity and then divided these patients into either a treatment group treated with 100,000 international units of vitamin D ₃ per month or an identical placebo group for a total of 12 treatment months. The psoriasis area and severity index (PASI) scores did not differ between the groups at any time point. Therefore, a direct cause of benefit of vitamin D ₃ supplementation for psoriasis could not be concluded, and so any benefit a patient experiences in their psoriasis disease severity from vitamin D ₃ therapy is likely because of chance (or placebo effect) alone.

If interventional studies show no significance whereas observational studies show significance, then disease B causes A.

The same study mentioned above, comparing treatment of vitamin D ₃ versus placebo on psoriasis, found a significant inverse relationship between PASI severity and 25-hydroxyvitamin D ₃ levels; however, replacing vitamin D ₃ provided no benefit. We can therefore assume that a low vitamin D level is a result of psoriasis disease activity, such that repletion has minimal to no role in treatment.