経験頻度、主観確率

最近、実感するのは、英語が日本語に比べ、総論と各論を明確に区別する言語であるということ。それが、冠詞や数に反映されているのではなかろうか。しかし、臨床的な推論において、一般論は、個人的な経験の回数などたかが知れているので、「３の法則」持ち出すまでもなく、"Always"や"Never"の極論の主張は限定的なものとなり、蓋然性を伴うものになる。

Always(33) 100%

• Endocarditis is always a concern in an elderly patient with a prolonged, nonlocalizing fever. "Ring around the Diagnosis"

Usually(145) 90% / Generally(69) 75%

• Erythema nodosum is usually a clinical diagnosis and does not generally require biopsy unless atypical features are present. "Footprints"

Often(198) / Frequently(64) 60%

• The serum phosphate level is often low in primary hyperparathyroidism, but the presence of renal failure in this patient clearly limits its usefulness. "Stopping Short of Certainty"
• Trichinosis frequently produces subclinical infections, but late presentations often include muscle aches, cardiac symptoms, and eye disturbances that are absent in this patient. "Where Are You From?"

Sometimes(41) 50%

• An atrial myxoma can sometimes cause episodic dyspnea, but usually not chest pain. "Rare × Rare"

Occasionally(28) 40%

• Back pain is occasionally the presenting symptom of endocarditis, although it is not usually associated with radiographic abnormalities. "Where Did Good Old Clinical Diagnosis Go?"

Rarely(39) 20%

• Bacterial infections are rarely a cause of chronic diarrhea. "Needle in a Haystack"

Hardly ever(0) 10%

Never(11) 0%

• Treatment with cephalosporins often results in a positive Coombs' test but almost never causes hemolysis. "Chest Pain with a Surprising Course"

参考

• オッズの非数値表現
• 医療における頻度を表す言葉と確率の接点

Say, Speak, Talk, Tellの違い

AllAboutに「Say, Speak, Talk, Tellの違い」という記事があった。結論は、Sayは「口から単語を出す」、speakは「一方通行に話す」、talkは「みんなで話し合う」、tellは「情報の伝達」だそうです。いづれの語もCPSでの登場頻度は、多くはない。そのなかで、"tell"が無生物主語の形で「情報の伝達」を意味したり、受身の形で「情報の有無」を表したりすることが多い。例文を下に引用しておく。

• Statistically, I would have to say that it is most likely that this man has substantial narrowing of at least one of his coronary arteries. "The Appropriate Degree of Diagnostic Certainty"
• In addition, the acute alteration in her ability to speak is more common with a stroke or a seizure than with a toxic or metabolic cause. "All in the Family"
• But her HIV infection is likely to be an issue when you talk to a cardiac surgeon. "Risky Business"
• We are told nothing in terms of medications that could raise the serum potassium level. "Things Are Seldom What They Seem"

オッズの非数値表現

オッズを言葉で表現する方法をまとめてみた。動詞、形容詞、助動詞、フレーズなど種々のものを集めて、４群に分類した。(　)内には、CPSコーパスでの使用例数を付した。

殆ど確定

• diagnostic of (13) 98.4% 
• certain (118) 95%

オッズ3以上

• indicate (219), indicative of (7) 92.1%
• can, almost certain (2) 90%
• very likely (3) 85%
• in keeping with (5) 82.1%
• probable (26) 75%

オッズ1/3〜３

• compatible with (48) 74.3%
• suggest (705) 69.8%
• suggestive of (105) 69.8%
• likely (654) 60%
• raise the possibility of (96) 52.9%
• frequent (24), could, may 50%
• might 40%

オッズ1/3以下

• possible 25%
• unlikely 15%
• remote 10%
• improbable 10%

同様の表現で、参照確率の資料がないもの。

associated with (391)、consistent with (306)、show (112)、argue againt (88)、represent (69)、point to (47)、characteristic of (44)、favor (28)、support for (15)、suspicious of (3)

参考文献

• 医療における頻度を表す言葉と確率の接点 
• The Interpretation of Phrases Used to Describe Uncertainty in Pathology Reports Malcolm Galloway and Taj Taiyeb Pathology Research International Volume 2011 
• Phraseology in pathology reports. A comparative study of interpretation among pathologists and surgeons R L Attanoos, A D Bull, A G Douglas-Jones, L J Fligelstone, D Semararo I Clin Pathol 1996;49:79-81
• How Medical Professionals Evaluate Expressions of Probability Augustine Kong, A.M., G. Octo Barnett, M.D., Frederick Mosteller, Ph.D., and Cleo Youtz, M.S. N Engl J Med 1986; 315:740-744

"worry" と "concern"

いづれもCPSのResponseで多く出てくる語。なんとなく臨床推論には相応しくないような感情的な意味合いを含む語だが、どのように違うのだろうか？
CSI シーズン５にSaraとGilの会話に次のようなものがある。

Sara: I don't have a death wish and I'm not a drunk in case you were worried.
Gil: I'm not worried. I'm concerned.

敢えて訳すと、前者が古英語由来で、後者が中世フランス語由来だから「心配はしてないが、危惧してるんだ。」ってなかんじになるのだろうか。ネイティヴの意見を聴くのが手っ取り早いかなと、ググルと "Worry is problem oriented; Concern is solution oriented" なんて記事があった。２つの語を対比させて箇条書きにされた部分を引くと、

Worry distracts us; Concern focuses us.
Worry disables planning; Concern helps us plan.
Worry blurs our vision; Concern clarifies our purpose.
Worry tends to give up; Concern perseveres.
Worry exaggerates; Concern pinpoints problems.
Worry focuses on self; Concern cares for others.

とあり、"worry"が感情的で、"cocern"は理性的な感じなんですかね、"Women worry more than men, but social concerns shared"なんて記事もあるようですし、"concern"を使うのが無難なのかも。CSPのコーパスでのコロケーションを見ると、いずれの語もaboutを介してpossibility、infection、diseaseなどの共通の語を支配するので、運用上はあまり違いがないようだ。実際、上記のGilの言葉に、Saraは、"Isn't that kind of the same thing?"と答えているのでした。差が余り無いのであれば、微妙なニュアンスの違いは無視して、いずれも「危惧」の意味だと覚えてしまうのが、「心配」だとか、「懸念」だとか使い分けなくてもいいし、訳の音節数も短いので思考のRAM容量の節約に貢献するかもしれない。

冠名疾患1ダース＋α

1. Crohn's disease

• Crohn's disease affecting the ileum, jejunum, and stomach can also result in vomiting and diarrhea. "A Rash Hypothesis"

2. Adult-onset Still's disease

• A truncal rash in a patient with a hectic fever (i.e., characterized by a daily spike in the temperature) suggests the possibility of adult Still's disease, but this is a diagnosis of exclusion. "Diagnosis Still in Question"

3. Whipple's disease

• Whipple's disease can present with fever and mood or memory disturbances. "Keeping an Open Mind"

4. Von Willebrand's disease

• The reduced factor VIII activity is consistent with — but not diagnostic of — von Willebrand's disease, since it is also seen in a state of factor VIII inhibition or deficiency. "A Bloody Mystery"

5. Wilson's disease

• Wilson's disease occurs in young adults and should be considered in any patient younger than 40 years of age with unexplained hepatitis. "A Sweet Source of Abdominal Pain"

6. Hodgkin's disease

• Clearly, a diagnosis of Hodgkin's disease should be considered in any patient with unexplained intrahepatic cholestasis, particularly a febrile cholestatic illness. "Empirically Incorrect"

7. Creutzfeldt-Jakob disease

• It is prudent to perform a limited workup to rule out conditions that may masquerade as Creutzfeldt–Jakob disease. "A Startling Decline"

8. Fabry's disease

• Patients with either Fabry's disease or PRKAG2 deficiency can present in adulthood with cardiac hypertrophy and preexcitation. "In the Thick of It"

9. Addison's disease

• Any cutaneous hyperpigmentation should at least raise the question of Addison's disease. "Impatient Inpatient Care"

10. Gaucher's disease

• Gaucher's disease can be diagnosed in adults, but the deposition of glucocerebroside in the liver, spleen, and bone marrow characteristically causes not only hepatomegaly but also splenomegaly, anemia, and thrombocytopenia. "A Sweet Source of Abdominal Pain"

11. Behçet's disease

• The history of oral ulcers is intriguing, but no other clinical findings suggest systemic lupus erythematosus or Behçet's disease. "On the Threshold — A Diagnosis of Exclusion"

12. Paget's disease

• Common metabolic bone diseases include osteoporosis, osteomalacia, Paget's disease, and osteitis fibrosa cystica due to hyperparathyroidism. "A Question Well Put"

プラスアルファとして、日本人の冠名疾患２つ。1961年に日本赤十字社の小児科医・川崎富作が患者を発見し、1967年に報告し名づけられた疾患と1972年に福岡大学病理学の菊池昌弘教授が報告した疾患。

Kawasaki's disease

• Spontaneous coronary-artery dissection may also occur in patients with underlying atherosclerotic plaque, connective-tissue diseases, immunological diseases or Kawasaki's disease. "Double Jeopardy"

Kikuchi's disease

• Posterior cervical lymphadenopathy, as seen in this patient, may occur with eczema of the scalp (which drains into the posterior cervical lymph-node chain), infectious mononucleosis or mononucleosis-like syndromes, human immunodeficiency virus (HIV) infection, tuberculosis, cancer, or, in young women, Kikuchi's disease (a self-limited, necrotizing lymphadenitis affecting the cervical lymph-node chains). "Venting the Spleen"

十大冠名症候群

1. Goodpasture's syndrome

• The absence of antineutrophil cytoplasmic antibodies argues against a diagnosis of Wegener's granulomatosis, and the absence of anti–glomerular basement membrane antibodies argues against a diagnosis of Goodpasture's syndrome. "Anatomy of a Diagnosis"

2. Sjögren's syndrome

• In light of the association between Sjögren's syndrome and lymphoma, the 30-kg weight loss, and the mild lymphadenopathy, lymphoma should be ruled out. "A Fractured Diagnosis"

3. Churg-Strauss syndrome

• Although asthma itself rarely results in peripheral-blood eosinophil counts of more than 800 cells per cubic millimeter, a possibility is Churg–Strauss syndrome, a condition characterized by asthma, eosinophilia, and sinus disease. "The Writing on the Wall"

4. Cushing's syndrome

• The elevated glucose concentration is probably due to diabetes mellitus, but there could be competing causes of hyperglycemia, including Cushing's disease or Cushing's syndrome, pancreatic disease, pheochromocytoma, and any number of drugs. "A Square Peg in a Round Hole"

5. Ehlers-Danlos syndrome

• The Ehlers–Danlos syndrome or Marfan's syndrome is less likely because of the patient's relatively normal stature and the absence of joint hypermobility and heart murmur. "A Fragile Balance"

6. Behçet's syndrome

• Behcet's syndrome would explain many of the patient's signs and symptoms, but the absence of ocular lesions or genital ulcers makes this diagnosis less plausible. "Saved by a Test Result"

7. Miller-Fisher syndrome

• I remain concerned about the possibility of the Miller Fisher syndrome or botulism and would also consider myasthenia gravis. "The Eyes Have It"

8. Marfan's syndrome

• There are patients with a so-called forme fruste of Marfan's syndrome who may present with only the cardiac abnormalities. "The Heart of the Matter"

9. Budd-Chiari syndrome

• The development of worsening abdominal pain, diarrhea, and hematochezia suggests the occurrence of mesenteric venous insufficiency, again raising the possibility of the Budd–Chiari syndrome. "A Window of Opportunity"

10. Fanconi syndrome

• The presence of glycosuria and aminoaciduria would support the diagnosis of Fanconi's syndrome. "A Fractured Diagnosis"

THE DIAGNOSTIC PROCESS

最近、臨床推論におけるムーブについて考えていたが、何も車輪の再発明をするまでもなく、右掲本の冒頭に「診断のプロセス」についての要約があることに、いまさらながら気がついた。初版本は読んでいたが、最近の認知心理学の進歩なども取り入れた記述に改訂されているので、買いの一冊。日本語に拘らなければ、kindle版の"Symptom to Diagnosis: An Evidence Based Guide, Second Edition (LANGE Clinical Medicine) " が、45%引きで手に入る。該当部分を引用する。

Constructing a differential diagnosis, choosing diagnostic tests, and interpreting the results are key skills for all physicians and are some of the primary new skills medical students begin to learn during their third year. The diagnostic process, often called clinical reasoning, is complex, but it can be broken down into a series of steps, diagrammed in Figure 1–1.
Step 1: Data AcquisitionData you acquire through your history and physical exam, sometimes accompanied by preliminary laboratory tests, form the basis for your initial diagnostic reasoning. Your reasoning will be faulty unless you start with accurate data, so the prerequisite for obtaining valid data is well developed interviewing and physical examination skills.
Step 2: Accurate Problem RepresentationThis step consists of developing a “problem synthesis statement,” a concise, single sentence summary of the main clinical problem and its associated context.
Clinical problems are symptoms, physical findings, test abnormalities, or health conditions for which diagnostic evaluation could be undertaken. The problem synthesis statement is meant to focus on the patient's most important problem, usually the chief complaint.
Context refers to pivotal points, generally one of a pair of opposing descriptors used to compare and contrast diagnoses or clinical characteristics; for example, old versus new headache, unilateral versus bilateral edema, smoker versus nonsmoker. Extracting pivotal points from the history and physical exam enables the clinician to focus a broad differential diagnosis to a more limited set of diagnoses pertinent to that particular patient. The prerequisite for being able to construct an accurate problem representation is knowledge of the pivotal points for specific clinical problems.
Step 3: Develop a Complete, Framed Differential DiagnosisThe process for developing a differential diagnosis will be discussed later in this chapter; subsequent chapters will present comprehensive, framed differential diagnoses specific for each problem discussed.
Step 4: Prioritize the Differential DiagnosisNot all diagnoses in a given differential are equally likely, or equally important. In order to effectively select diagnostic tests and therapies, it is necessary to select a “leading hypothesis,” a “must not miss” hypothesis, and other “active alternative hypotheses” (see full discussion later). The prerequisites for this step are knowledge of pivotal points; typical or “textbook” presentations of disease; the variability of disease presentation; and which diseases are life-threatening, very common, or easily treatable. It is also necessary to know how to estimate pretest probability, and which history, physical, or laboratory findings are so specific for a disease they are diagnostic; in other words, such findings are “fingerprints” for the disease.
Step 5: Test Your HypothesisSometimes you are certain about the diagnosis based on the initial data and proceed to treatment. Most of the time, however, you require additional data to confirm your diagnostic hypotheses; in other words, you need to order diagnostic tests. Whenever you do so, you should understand how much the test will change the probability the patient has the disease in question. The prerequisite for this step is knowing the sensitivity, specificity and likelihood ratios (LRs) of the tests you have chosen, knowing how to interpret these test characteristics, and understanding how to determine posttest probability using pretest probabilities and LRs.
Step 6: Review and Reprioritize the Differential DiagnosisRemember, ruling out a disease is usually not enough; you must also determine the cause of the patient's symptom. For example, you may have eliminated myocardial infarction (MI) as a cause of chest pain, but you still need to determine whether the pain is due to reflux or muscle strain, etc. Whenever you have not made a diagnosis, or when you encounter data that conflict with your original hypotheses, go back to the complete differential diagnosis and reprioritize it, taking the new data into consideration. Failure to carry out this step is one of the most common diagnostic errors made by clinicians and is called “premature closure.”
Step 7: Test the New HypothesesRepeat the process until a diagnosis is reached.