How to Write the Discussion

The Hardest Part of the Paper

The true meaning of the data

obscured

A faulty discussion presented

Common Negligence

• Not discussed adequately.

• Engrossed with the trees(the data) without noticing how much sunshine had appeared in the forest.

A Faulty Discussion

“Not with a bang(突然见效或成功 ) but a whimper(呜咽声 ).”

—T. S. Eliot

A Faulty Discussion

“Many a paper loses much of its effect because the clear stream of the discussion ends in a swampy(沼泽的 ) delta (三角州 ).” —Anderson and Thistle

A Good Discussion

“Finally, good writing, like good music, has a fitting climax(高潮 ).”

— Anderson and Thistle

Defining Scientific Truth

• Not necessary to reach cosmic conclusions (the whole truth).

• Shine a spotlight on one area of the truth.

• Describe the meaning of your little bit of truth simply (The simplest statements evoke the most wisdom).

• Leave the “whole truth” to the ignoramuses(不学无术的人 ).

The Primary Purpose of the Discussion

To show the relationships

among observed facts.

The Process of the Job

• From the presentation to the discussion of the results.

• Not simply recapitulating(概括 ) but restating the results in a meaningful way.

The Coverage of the Process

1. Previous investigations2. Intentions3. Results4. Findings5. Agreement(s)6. Contrast(s)7. Indications8. Limitations9. Implications10.Conclusion(s)11.Further investigation

The Contrast between the Results and

Discussion

• Factual vs. Explanatory• Objective vs. Interpretative • Numerical vs. Analytical• Informative vs. Critical• Presentable vs. Evaluative• Valid vs. Illuminating

Guidelines in General

• Describe the results of the current study in the past tense and the known or proved facts in the present tense.

Guidelines in General

• Do not simply repeat what you have already said in the Results Section; try to present the principles, relationships, generalizations shown by the results obtained.

Guidelines in General

• Assess the validity of the results.

Guidelines in General

• Show that your results and interpretations agree (or contrast) with the previously published work.

Guidelines in General

• Discuss any results that do not “fit in”; do not try to hide them or to “fudge (捏造 )” the data.

Guidelines in General

• Explain the significance of the results; comment on their theoretical and practical implications.

Guidelines in General

• State your conclusions as clearly as possible, and summarize your evidence for each one.

Guidelines in General

• Point out possible lines of further investigation.

LET’S READ AN EXAMPLE AND

BE CRITICAL.

Chronic Daily Headache: Identification of Factors Ass

ociated with Induction and Transformation

Marcelo E.Bigal. MD, Phd; Fred D. Sheftell, MD; Alan M. Rapoport, MD; Stewart J. Tepper, MD;

Richard B. Lipton, MD

Chronic daily headache (CDHacrinym) is one of the more frequently encountered headache syndromes at major tertiary care centers.1-3 Its prevalence is almost 5% in the general population, and it is the most common headache problem seen in clinics specifically devoted to headache management.4-6 Most patients with CDH report their role functioning and well-being to be frequently and severely impaired, with a consequent reduction in quality of life.7.8

It is clear that comorbid factors play a significant role in the development and maintenance of headache that occurs on a daily or near-daily basis.9.10 Originally coined by Feinstein,11 the term comorbidity is used to refer to the greater than coincidental association of two conditions in the same individual.12

Psychiatric issues are prominent among patients with more difficult headache syndromes, particularly CDH. Studies have shown that depression occurs in 80% of patients with chronic migraine (CM), the most common type of CDH.13.14 Compared with controls, stressful life events such as divorce, widowhood, separation, and problems with children are more likely to occur in individuals with CDH.15 Assessment of psychopathology shows a characteristic profile of hypochondria(疑病症 ), depression, and hysteria in patients with CHD (with the revised version of the Minnesota Multiphasic Personality Inventory 2).16

Potential somatic(躯体的 ) comorbidity in CDH has received less study than the psychological aspects. Hypertension,17-20 alcohol overuse,21 and sleep disturbances have been reported to be associated with CDH.22 What factors specifically contribute to the transformation from episodic into CM and to the de novo (再次 ) development of new daily persistent headache(NDPH) remain largely unknown.

The purpose of this study was to identify somatic factors and behaviors associated with the development of CM and NDPH.

METHODS The clinical records and headache diaries of 803 randomly selected patients seen between 1980 and 2001 at the New England Center for Headache were reviewed. Cases were included if they fulfilled the following criteria: daily or near-daily headache lasting more than 4 hours if not treated, occurring more than 15days/month, fulfilling the Proposed Headache Classification for Chronic Daily Headache described by Silberstein and Lipton,23 headache fulfilling the criteria for chronic posttraumatic headache(CPTH) of the International Headache Society(HIS) Classification of Head and Facial Pain;24 headache fulfilling the IHS criteria for migraine;24 and exclusion of the disorders listed in groups 5 through 11 of the IHS classification system, with the exception of CPTH.

During the entire time period from which the cases were extracted, the clinic used a uniform clinical intake form and headache calendars.25.26 A questionnaire was mailed to patients prior to their initial consultation, to be completed and returned at the time of the visit, when it would be reviewed and checked by a nurse and by the headache specialist. This questionnaire contained several multiple-choice and fill-in-the-blank items relating to past history, triggers, and patient behavior.

We used a randomized, case-control design to study the following case groups: 1) CM with analgesic overuse (ARH), n=399; 2) CM without analgesic overuse, n=189; and 3) NDPH, n=69. These “active” groups were compared with two control groups: 1) migraine, n=100; and 2) CPTH, n=65.

Associated medical conditions, chronic and recently acquired, were assessed, and we investigated the case groups for any association with the following somatic or environmental factors: 1) disorders due to hypersensitivity, including allergic rhinitis(鼻炎 ), urticaria(风疹 ), and drug hypersensitivity, but excluding asthma; 2) asthma; 3) hypothyroidism(甲状腺机能减退 ); 4) hypertension; 5) chronic pain syndromes involving pain other than the headaches; 6) consumption of alcohol more than 100 mg of caffeine daily; 7) consumption of alcohol more than three times a week; and 8) daily use of tobacco.

We used descriptive statistics and the chi-square test for gender comparisons. We used analysis of variance with posttest for age comparisons. We analyzed comorbidities by the two-sided Fischer’s exact test, and we calculated the odds ratio considering a 95% confidence interval, using the approximation of Woolf.

RESULTS

Table 1 displays the characteristics of each group. There were no significant differences concerning in gender ratio groups. One case group (ARH) and one control group had a mean age significantly higher than other groups.

Table 2 summarizes the odds ratio, confidence interval, and the P value obtained from the comparison among the “active” groups and the episodic(偶尔发生的 ) migraine group. Several strong correlations are evident, with odds ratios as high as 16.0 (for NDPH and hypothyroidism 甲状腺机能减退 ). The following associations were identified: 1) ARH: hypertension and daily consumption of caffeine; 2) CM: allergies, asthma, hypothyroidism, hypertension, and daily consumption of caffeine; and 3) NDPH: allergies, asthma, hypothyroidism, and consumption of alcohol more than three times per week.

Results of the comparison between the “active” groups and CPTH are shown in Table 3, and several strong correlations are again evident. The highest odds ratio was that observed for NDPH and hypothyroidism (10.3 with a 95% confidence interval ranging from 2.3 and 46.7). The following associations were found: 1) ARH: asthma and hypertension; 2) CM: allergies, asthma, hypothyroidism, hypertension, and daily consumption of caffeine; and 3) NDPH: allergies, asthma, hypothyroidism, and consumption of alcohol more than three times per week. The association found for NDPH and CM with CPTH as a control group were the same as those found when the episodic migraine group was used as a control. Hypertension and ARH were associated with ARH only with CPTH as a control, and consumption of caffeine was associated only with episodic migraine as a control.

Discussion Failure to identify comorbid medical conditions and behaviors can create misleading medical statistics and may cause spurious(不合逻辑的 ) comparisons during the planning and assessment of medical treatment2

7. Comorbid disorders may alter the clinical course of illness, affecting the time of detection and treatment. Mechanisms contributing to comorbidity include that following:27,29 a) a relationship by pure chance only; b) one disorder causes the other (eg, diabetes and diabetic neuropathy); c) environmental risks shared by the conditions; d) genetic risk factors shared by the conditions (ie, there is a common biology underlying both conditions ).

Although the pathophysiology of pain in primary CDH is unknown and may be dependent on the clinical subtype, several biologic mechanisms may underlie the process30-33: 1) prolonged or heightened excitation of peripheral nociceptive (疼痛的 ) afferent (传入的 ) fibers (perhaps due to neurogenic(神经性的 ) inflammation); 2) enhanced responsiveness of the nucleus caudalis and dorsal horn neurons (central sensitization); 3) altered modulation of central pain; 4) spontaneous central pain due activation of the “on cells” in the medulla; or 5) a combination of these. Each mechanism can presumably be catalyzed by the overuse of analgesics and other acute care medications.

Comorbid factors also may be significant contributors to the development and maintenance of headaches that occur on a daily or near-daily basis. Psychological comorbidity is common in patients with CDH, and stress has been identified as a trigger of the transformation process33. The potential role of somatic (肉体的 ) comorbidity in CDH has received less study than the psychological aspects. Examining the results of our CHD versus episodic migraine comparison alone, one could conclude that there exists an association between CHD and various somatic conditions, but not that a causal relationship necessarily exists. The comorbidities identified from that comparison alone simply represent a consequence of chronic headache unrelated to the development of CHD.

This was our rationale (基本原理 ) for using a second control group of patients with CPTH. CPTH can be considered a secondary form of CDH (particularly of NDPH ), with trauma being the triggering factor related to the development of the headache syndrome. Theoretically, the patient with CPTH may possess the same comorbidities consequent to a chronic pain syndrome as patients with CDH, but in the case of CPTH, these comorbidities play no causal role. We thus assumed that our use of episodic migraine as a control would allow us to identify comorbidities relatively unique to CDH, and while using CPTH as a control, we could identify comorbidities related more specifically to transformation or to de novo (再次 ) development of NDPH.

We found two basic patterns of association: ARH with few detected comorbidities; and CM and NDPH with multiple, strong, and consistent associations. Hypertension was associated with ARH and CM. Mathew et al34 reported that patients with CDH evolving from initially episodic migraine were more likely to be hypertensive than patients with episodic headache. Other investigators similarly have found CHD and hypertension to be associated.35-37

Allergies and asthma were associated with both CM and NDPH. The association between migraine and allergic conditions (including food allergies, asthma, hay fever, and bronchitis) has been studied previously38,39. In a case-control study, Bille40reported that children with migraine had twice the risk of allergies than those without migraine. Migraine frequency may decrease in patients with comorbid asthma who take montelukast, a specific anti-asthmatic medication.41 Even with allergies and asthma being more common in migraineurs than nonmigraineurs (odd ratio=1.7) 27, our data indicate that asthma and allergies may be much more common in CM and NDPH than in migraine, and this is the first time that this association has been demonstrated in a CDH population. 27

A strong correlation was found between hypothyroidism(甲状腺机能减退 ) and NDPH (odds ratio = 16.0 when compared with migraine controls and 10.3 when compared with CPTH) and also with CM (odds ratios of 8.4 and 5.4, respectively). An association between headache and hypothyroidism rarely has been reported.42,43 Moreau et al44 reported that 30% of patients with hypothyroidism described their headaches improving after initiation of thyroid hormone replacement. Our study is the first to suggest a possible role for hypothyroidism in the development and maintenance of CDH.

Concerning patients’ behavior, we found two associations: 1) CM and daily consumption of caffeine, and 2) NDPH and alcohol consumption. Caffeine overuse may contribute to the development of CDH, and the headache syndrome improves after its withdrawal.” Although the relationship between alcohol consumption and CDH is poorly described, alcohol is a well-known trigger of migraine and cluster headache, and alcohol-induced headache may follow alcohol ingestion.24 Mathew has reported a relationship between alcohol overuse and CDH. 46

We believe our data may have clinical relevance. Several strong correlations were identified, some not previously recognized, and some may have a causal relationship to CDH. Thus, transformation of previously episodic headache or development of NDPH may be related to medical conditions and behaviors in addition to the frequently incriminated precipitant overuse of analgesic. These will require confirmation by prospective studies.

Assignment

• Read your target journal;• Find a well-written Discussion;• Indicate, in a particular way, which is

which and what is being discussed;• Make a list of stock phrases;• Make your homework look appealing;• Submit your assignment to your

instructor at the oral test.