The process of asking and answering questions is summarized in the “Information Pathway” shown below:
Each step in the pathway also has inputs that can help or hinder progression to the next step, such as skills, the environment, culture, and attitude.
The first step in the information pathway is to recognize that an information need exists. The willingness to recognize that one’s knowledge might not be complete would appear to be driven primarily by physician attitude and personality. Work by educators in the area of critical reflection suggests that this skill is mature and high-level,[13] but that it might be undermined by an emphasis on what Argyris[15] calls single-loop rather than double-loop learning. Consider the example of a patient who has a cough, which is diagnosed as bronchitis, and who is given erythromycin. When the patient returns 2 days later with persistent symptoms, the physician changes the antibiotic. This behavior is single-loop learning, because the physician reacts to the situation without questioning the underlying assumptions. A more reflective physician would practice double-loop learning: when the patient returns, both the underlying diagnosis and the treatment strategy are questioned. Maybe the patient does not have bronchitis. Perhaps the patient has reflux or allergies. What if the bronchitis is viral, and the antibiotic is not going to help anyway? Do antibiotics help even if the bronchitis is bacterial? The reflective physician, even in this simple clinical situation, generates a host of important questions.
In addition to generating more clinical questions, the reflective physician is more likely to practice patient-centered medicine, tailoring the management strategy to the patient’s clinical picture, needs, and situation. For example, taking a one-size-fits-all approach to the management of sore throat by ordering a rapid antigen test for everyone who has a sore throat provides too much care for some and not enough for others. By reflecting on practice and asking whether a rapid antigen test is needed for a particular patient, care is individualized and outcomes are perhaps improved.
Unfortunately, traditional medical education emphasizes knowing the right answer more than asking the right question. Too often, asking a question is punished by giving the busy intern or medical student the task of reporting the answer back to the group the next day! The traditional paternalistic view of physicians as all-knowing might make them afraid to admit that they do not know everything, and less likely to appraise critically their current practices. Such phrases as “This is the Michigan State (or Harvard, or Michigan, or Duke) way,” and “We’ve always done it this way, and my patients do fine,” further undermine our willingness and ability to critically reflect on our practices. Argyris has described similar defensive strategies in the business setting, and called them learned incompetence. Just as General Motors was slow to change its practices during the 1970s and 1980s, physicians find many reasons to justify the status quo and resist positive change.
The decision to pursue a clinical question is perhaps the most complex step in the information pathway, influenced by physician attitude, personality, and work ethic; the characteristics of the question; and the practice environment. Previous research has shown that the decision to pursue a clinical question is driven most strongly by the importance of the question and the perceived availability of an answer to the question.[7] The former might vary depending on physician characteristics such as conscientiousness, work ethic, and sense of duty to the practice. For example, one physician might believe that it is important to find out the best way to treat corneal abrasions, because the wrong decision might adversely affect some patients. Another, however. might be more complacent and believe that using a patch has always worked well enough.
The perceived availability of an answer to the question depends in part on the physician’s knowledge of available resources. For example, a physician wondering about the efficacy of steroids in preterm labor might keep wondering because it would seem to be too difficult to find the answer. Knowing that the Cochrane Database of Systematic Reviews has many excellent systematic reviews on perinatal topics, however, might stimulate the physician to look there for an answer, especially if that physician had the necessary computer skills. Finally, environmental factors, such as access to information sources and the time available between patients, will affect the decision to pursue a clinical question. Having the Cochrane Database of Systematic Reviews abstracts on a handheld or office computer in the clinical area would eliminate a trip to the library and reduce the time needed to access the review.
The ability to satisfy an information need and answer a clinical question once the physician decides to pursue it is determined by the knowledge of available resources, skill at framing a question and searching these resources, and the context or environment. Physicians tend to frame questions in relation to specific patients [4,8] rather than in a generalizable fashion, as is advocated by teachers of evidence-based medicine.[16] For example, a family physician might ask, “I have this 38-year-old patient with dyspepsia, whose father was diagnosed with pancreatic cancer. He’s really worried that he has pancreatic cancer, and I want to reassure him. What is the best test to rule out pancreatic cancer for him? Should I order a sonogram?” A clinical epidemiologist or evidence-based medicine proponent might frame this question as, “What is the test with the best negative likelihood ratio for pancreatic cancer among outpatients like this with dyspepsia?” Finally, a research article might provide information in terms of sensitivity and specificity, but not calculate the likelihood ratio. In addition, the researchers might present data as false-positive and false-negative rates rather than sensitivity or specificity, further confusing the naive reader.
The decision to implement the answer to a clinical question in practice will depend on such environmental factors as cost, health system constraints, patient acceptance, acceptance of colleagues, and local practice patterns. The amount of work needed to implement a change as well as the importance of the question are both important. For example, it is easy to stop patching corneal abrasions but very difficult to establish a stroke unit in a local hospital. Nevertheless, if the outcomes are important enough in terms of reduced morbidity and mortality, then it should be worth the effort to establish such a unit.
Richard Smith, editor of the British Medical Journal, states: “New information tools are needed: they are likely to be electronic, portable, fast, easy to use, connected to both a large valid database of medical knowledge and the patient record, and a servant of the patient as well as doctors.”[4] As we learned in the introductory module, Shaughnessy and colleagues[20] have described the usefulness of medical information as follows:
|
Usefulness of |
= |
Relevance X Validity |
Thus, the most useful information is relevant to a practice, highly valid, and takes very little work to acquire. This insight can guide in defining and even designing resources for answering clinical questions at the point of care.
The traditional approach to evidence-based medicine teaches a five-step approach: question, search, appraise, apply, and evaluate. This approach emphasizes validity assessment more than relevance, advocates formal MEDLINE searches, and encourages physicians to read the original research literature and do their own critical appraisal.[21] This time-consuming process, however, is impractical for the busy clinician. Information mastery takes a more balanced approach and emphasizes an initial assessment of outcomes and relevance before proceeding to the assessment of validity.
The difference between traditional evidence-based medicine and information
mastery is shown graphically in Figure
2. Whereas the traditional approach remains the basis for those who perform
meta-analyses, systematic reviews, and critical appraisals for secondary
literature journals, information mastery provides a much more accessible and
efficient way for physicians to keep up-to-date and answer their clinical
questions rapidly at the point of care. For example, it is much easier to search
the Cochrane Library, Best Evidence, or InfoRetriever CD-ROMs
than all of MEDLINE. Also, rather than beginning with a time-consuming validity
assessment, Shaughnessy et al[20] advocate reading the abstract to
find out whether the outcomes are patient-oriented and whether the conclusion
recommends a change to your current practice. If either answer is no, you can
stop and move on to the next article.