10-1 PHYSICIANS� ATTITUDES ABOUT PRESCRIBING AND KNOWLEDGE OF THE COSTS OF COMMON MEDICINES.
      Physicians were predisposed to being cost-conscious in their prescribing habits, but lacked knowledge about actual costs. Considering the benefit/harm-cost ratio is the basic consideration of therapy. We often fail to consider the cost denominator. We should know the costs especially of drugs used long term (eg, hypertension, diabetes, lipid disorders).
      It is easy to access costs on the internet, for both generics and brand-names. The web pages of major pharmacies (eg, www.cvs.com 1-888-607-4287 ) post actual retail costs.

10-2 PULSE PRESSURE AND MORTALITY
      The best single measure of blood pressure to predict mortality in older people appears to be the pulse pressure. A high systolic combined with a low diastolic BP therefore correlates with increased risk.
      In middle-aged persons, systolic BP alone is a major risk factor.
      In older persons, the pulse pressure is the major risk factor.
      Systolic BP correlates closely with pulse pressure in older patients
      Thus, systolic BP alone can be considered a common denominator of risk at all ages.
      Treatment is aimed at reducing systolic BP

10-3 LOW-DOSE HYDROCHLOROTHIAZIDE AND PRESERVATION OF BONE MINERAL DENSITY IN OLDER ADULTS
      Low-dose hydrochlorothiazide given for 3 years preserved BMD. Low-dose thiazides are still considered the first choice anti-hypertensives � preservation of BMD is an added benefit.

10-4 THE DIAGNOSTIC VALUE OF HISTORICAL FEATURES IN PRIMARY HEADACHE SYNDROMES
      The most sensitive clinical features for migraine are nausea, phonophobia, and photophobia. Foods, especially cheese and chocolate, are triggers for migraine. Aura, especially visual, predominate in migraine and when present establish the diagnosis.
      TTHA, the most common HA, is recurring, pressing, and band-like with few migrainous features. Phonophobia, photophobia and nausea are not common. Stress and hunger are triggers.
      Cluster HA is a distinct syndrome.
      "A carefully taken history can establish the diagnosis of migraine, cluster or tension-type headache with sufficient confidence that no need will exist for additional evaluation."

10-5 PREVENTING INFLUENZA IN HEALTHY ADULTS: The Evolving Story
      Flu vaccine for healthy adults will prevent influenza in some individuals each year. The absolute number will vary with the severity of the epidemic and the match between the native influenza virus and the vaccine virus.
      Healthy health care workers should take vaccine every year.

10-6 USE OF LIPID LOWERING DRUGS FOR PRIMARY PREVENTION OF CORONARY HEART DISEASE: Meta-Analysis of Randomized Trials
      Treatment with lipid controlling drugs over 5 to 7 years reduced coronary heart disease events in persons with no known cardiovascular disease. (Primary prevention)
      National Guidelines have suggested starting primary prevention treatment in persons with a 3% annual risk of CHD events. But the decision to take drugs depends to a large extent on personal preference. Clinicians should assess risks in each patient and if they are considered at higher risk than average, should inform them of the benefit/harm-cost ratio to guide their individual decision. Statin drugs are costly. RTJ

10-7 STATINS: UNDERUSED BY THOSE WHO WOULD BENEFIT
      Statins are effective, easy to take, and relatively safe (and very expensive). Many people who would benefit from taking them are not doing so. Underuse is most apparent in the secondary prevention of heart disease in patients with known atherosclerotic disease for whom there is overwhelming evidence that statins are highly beneficial. Treatment should start as soon as a clinical diagnosis is made.
      Undertreatment is also a problem for the larger population of people who do not have manifest atherosclerotic disease (primary prevention). "There is no longer any doubt that statin treatment benefits those who are at substantial coronary risk1 but do not have established atherosclerotic disease.
      In practice, clinicians might consider treatment for those of younger age who have strong risk factors such as familial hypercholesterolemia or diabetes. Problems include cost and issues of long-term safety, although statins are reasonably safe and less likely to cause serious harm than aspirin.
      For primary prevention of persons at higher than average risk of atherosclerotic events, MDs should explain the benefit/harm-cost ratio to each individual and allow him/her to make a personal informed choice.

10-8 CAN WE IMPROVE DIAGNOSIS OF ACUTE APPENDICITIS?
      Given the false positive (12%) and false negative rate (5%) of ultrasound, should ultrasound be allowed to override clinical judgement? There is probably no role for ultrasound where clinical evidence is convincing.
      The main role of ultrasound may be for the equivocal case, where a combination of repeated clinical assessment and ultrasound may provide the additional information required to determine whether surgery is necessary. Patients should not be sent home after negative ultrasonography unless there are also clinical grounds for their discharge.

10-9 THE PRACTICAL MANAGEMENT OF CLAUDICATION
As A Marker For Cardiovascular Disease It Needs Active Treatment
      Since peripheral vascular disease is a strong risk factor for myocardial infarction and stroke, it might be expected that patients would benefit by reducing risk of MI and stroke.
      Claudication is a marker for generalized atherosclerotic disease. We should look at the patient�s overall vascular risk and act accordingly. Preventive interventions include: smoking cessation; exercise; lipid control; and antiplatelet therapy.

10-10 A COMPARISON OF GLYBURIDE AND INSULIN IN WOMEN WITH GESTATIONAL DIABETES MELLITUS
      Gyburide (DiaBeta; generic) is a clinically effective alternative to insulin in women with gestational diabetes. It is as safe as insulin.

10-11 SCREENING STRATEGIES FOR EARLY DETECTION OF LUNG CANCER
      The editorialist argues that with CT scanning and improved cytology, many lung cancers can be diagnosed at an early stage and cured.
      "Now is the time to screen for early-stage lung cancer." Screening heavy smokers with airflow obstruction with sputum cytology testing for central lesions and CT for peripheral lesions can identify and harvest the "low hanging fruit".

10-12 ROUTINE SCREENING FOR LUNG CANCER?
Lead Time Bias; Length Time Bias; Contamination Of The Control Population; Overdiagnosis
      There is not enough data to predict the benefit/harm-cost ratio of routine screening for LC. This commentator argues that screening is not of proven benefit. There are too many biases in the way of proof.
      On an individual basis, patients at high risk (long-time smokers with decreased FEV1) may be made aware of the benefits, harms, and costs of screening, and be given the opportunity to choose for themselves.
      See the original for a definition of lead-time bias, length-time bias, overtreatment, and contamination of the control population.

10-13 CHOLESTEROL EMBOLI SYNDROME
      Cholesterol embolization (CE) is a common complication of arteriography, vascular surgery, thrombolysis, and anticoagulation in elderly patients. The diagnosis is not often considered.
      CE is a serious multisystem disorder, more common than has been supposed. It is characterized by a classic triad: 1) livido reticularis, 2) acute renal failure, and 3) eosinophilia
      CE should feature strongly in the differential diagnosis of acute renal failure in elderly patients after angiography or other interventions. Clinicians should be aware of the delayed onset which often obscures the causative link. It may, however, occur without instrumentation.

10-14 UPRIGHT POSTURE AND POSTPRANDIAL HYPOTENSION IN ELDERLY PERSONS
      Meal ingestion was significantly related to a fall in systolic BP in elderly persons. After the meal, 22% had resultant symptomatic hypotension (systolic < 80 mm Hg) vs 12% when tested before the meal.
      The problem of post-prandial hypotension and syncope leading to falls is common in nursing home patients. We should take care to keep individuals with this risk sedentary or recumbent for a time after meals, not allowing them to rise quickly.

10-15 MELATONIN, CIRCADIAN RHYTHMS, AND SLEEP
      Disturbances in circadian rhythms often result in disturbances in sleep. Some examples; jet lag; shift work; and the sleep disorders that occur in totally blind persons with free-running circadian rhythms (ie, rhythms that are not synchronized to the 24-hour day). The hormone melatonin can be used both to characterize and treat such disorders.

10-16 GRAVES' DISEASE
      I enjoyed this review. I abstracted some points I felt needed emphasis, and some I did not know or had forgotten. RTJ

10-17 ORAL CONTRACEPTIVES AND BREAST CANCER.
      There was a significant correlation between ever-use of OCs and risk of BC in sisters and daughters of probands (RR = 3.3 compared to marry-ins), but not among granddaughters or nieces. Women using OCs who had 3 family members with BC had a RR of 5; for those with 5 affected family members RR = 11. "These data offer strong support for the amplified effect of estrogen in the presence of genetic risk for BC."
      However, the increased risk was seen only in users before 1975 when formulations contained higher doses of estrogen and progestins. Among first degree relatives who used OCs after 1975, the RR of BC was 0.9 compared with non-users.
      First degree relatives of a woman with BC should use other contraceptive methods.

10-18 RELIGION, SPIRITUALITY, AND MEDICINE: Application To Clinical Practice
      Patients want to be seen and treated as whole persons, not as diseases. A whole person is someone whose being has physical, emotional, and spiritual dimensions. Ignoring any one of these aspects of humanity leaves the patient feeling incomplete, and may even interfere with healing. For many patients spirituality is an important part of wholeness.
      The editorialist quotes 4 simple questions the ACP suggests be asked of those with terminal illness which acknowledges the spiritual life of the patient and might be appropriate and helpful.

RECOMMENDED READING

10-18 RELIGION, SPIRITUALITY, AND MEDICINE: Application To Clinical Practice

REFEREENCE ARTICLES

10-4 THE DIAGNOSTIC VALUE OF HISTORICAL FEATURES IN PRIMARY HEADACHE SYNDROMES
10-16 GRAVES' DISEASE

1. Conducted a survey of internal medicine and general medicine clinicians. (134 responded; 70% house officers)
2. Questions concerned estimates of the average wholesale price of a 30-day supply of commonly used drugs.

1. The great majority considered costs an important consideration in the prescribing decision.
2. About 3/4 were willing to sacrifice some degree of effectiveness to make drugs more affordable.
3. Over 75% often were unaware of actual costs. Only 1/3 had easy access to drug cost information.
4. Almost all gave consideration to the cost when patients were self-paying. Only 30% considered this when patients had Medicaid or an HMO with a prescription plan.1 A few physicians preferred brand names over generics regardless of cost. Estimates were more accurate for generics than for brand-name drugs.
5. Estimates of a month�s supply of commonly used drugs were too low in 40% of cases; accurate in 45%. The costs of brand-name and expensive drugs were most likely to be underestimated.
6. Attending physicians were more cost-conscious than house officers.
7. Physicians rarely asked patients about the cost of the drugs.

1. Straining to pay the cost of their drugs is a common occurrence for the poor and elderly and contributes to suboptimal care. High costs may prevent patients from filling their prescriptions and may reduce compliance, especially in treatment of chronic diseases.
2. In general, physicians appeared predisposed to practice cost-effective medicine. But they were often ill-equipped with the requisite knowledge to do so. They were likely to underestimate costs of expensive drugs.

1. Measured BP in over 9000 persons age 65 to 102 in 1981.
2. Followed for an average of 10 years.

1. Over 10 years over 4500 subjects died; 2300 of cardiovascular causes.
2. The lowest overall death rate occurred in those with systolic less than 130, and diastolic 80 to 90.
3. The highest mortality occurred in those with a systolic 160 or more and diastolic less than 70.
4. Both low diastolic and elevated systolic independently predicted increases in cardiovascular and total mortality.
5. PP correlated strongly with systolic pressure. But, PP was a slightly stronger predictor of both cardiovascular and total mortality than systolic pressure.

1. "These prospective data indicate that, in older people, both (high) systolic and (low) diastolic pressure provide important and independent prognostic information about the risk of cardiovascular and total mortality."
2. As in middle-aged people, in older persons, higher systolic pressure predicts linear increases in risk, with little evidence of confounding by other comorbid conditions. There is a clear increase in risk among those with borderline high (140-159 mm Hg) systolic pressure.
3. Low diastolic pressure is also a marker of increased risk of death, independent of systolic pressure. This association is largely explained by the confounding effects of frailty and comorbid conditions.
4. Pulse pressure, although highly correlated with systolic pressure, has the advantage of incorporating the effects of both high systolic pressure and low diastolic pressure. Prognostic models including PP are simpler and have nearly the same predictive value as models including both systolic and diastolic pressure.
5. PP appears to be the single best BP measurement to predict mortality risk in the elderly.
6. In healthy, middle-aged populaitons, PP is not a consistent and independent risk factor for cardiovascular disease. (Ie, in younger persons, systolic is a better predictor than PP.)
7. Several studies have indicated that the prognostic significance of systolic pressure increases with age, while that of diastolic pressure and mean arterial pressure decreases.
8. Also, "Pulse pressure appears to be the best measure of BP to predict heart failure in the elderly."
9. Several plausible mechanisms: arterial stiffness increases afterload and myocardial work, impairs ventricular relaxation, and causes ischemia. PP is strongly correlated with left ventricular hypertrophy, a known risk factor for cardiovascular events. It is also correlated with the presence of atherosclerosis . The increase in shear stress and pulsatile strain may promote primary atheroma development, and may contribute to rupture of vulnerable plaques.
10. Borderline elevations of systolic pressure (140-159) were associated with significant elevations in risk, particularly among individuals with a low diastolic pressures.
11. Therapies of potential value in improving arterial compliance include ACE inhibitors, nitrates, and a low salt diet.

PP is a major risk factor in the elderly.
Systolic BP correlates with PP in the elderly.
Therefore, systolic BP alone may be considered a major risk factor at all ages.
Treatment should aim at reducing systolic RTJ

1. Randomized, double-blind, placebo-controlled trial followed over 300 healthy normotensive adults age 60 to 79. (Mean = 68)
2. All received increased calcium intake.
3. Randomized to: 1) 12.5 mg hydrochlorothiazide daily, 2) 25 mg hydrochlorothiazide daily, or 3) placebo.
4. Measured BMD by X-ray absorptiometry at baseline and at 3 years.
5. Follow-up = 3 years.

1. At 3 years, differences in hip bone BMD compared with placebo: + 0.8 percentage points for 12.5 mg; +0.9 percentage points for 25 mg.
2. Percentage change in the spine was significantly greater in the 25 mg group (+0.8%) compared with placebo.
3. No significant differences in total BMD.
4. Treatment differences were greater in women than in men.

1. There was a dose response relationship to hydrochlorothiazide over 3 years.
2. Low-dose hydrochlorothiazide in healthy elderly subjects prevented bone loss at the hip and spine. At 3 years, the 25 mg hydrochlorothiazide group gained about 1% in BMD when compared with placebo.
3. The mechanism is likely the action of thiazides on the kidney tubule to decrease calcium excretion.
4. Despite accumulating evidence that low-dose thiazides reduce risk for stroke, congestive heart failure, coronary disease, and total mortality among patients with high BP, use has declined in recent years. �The benefits of thiazide on bone should be considered in decisions about long-term pharmacologic therapy for high blood pressure. �
5. If preservation of BMD accumulated over 10 to 20 years, the modest effects observed in this trial could possibly explain the one-third reduction in hip fracture observed in many observational studies.

1. Searched the literature concerning the sensitivity of clinical and historical features for the diagnosis of migraine, tension-type, and cluster HA. Classified them according to the ISH criteria as the standard for diagnosis.
2. Clinical features include pain character, location, onset, precipitants, aura, other associated symptoms, duration, frequency, and time-course.
3. Calculated sensitivities and specificities of particular clinical features with a gold standard described by the IHS.

1. For migraine, the most frequent and specific features are: 1) nausea, 2) exacerbation by physical activity, 3) photophobia, 4) phonophobia, and 5) aura.
2. Positive likelihood ratios:

   Nausea	Phonophobia	Photophobia
   19	5	6

   (Ie, in 20 patients with a headache accompanied by nausea, 19 will have migraine, and one will have a headache of other origin. RTJ )
   (Likelihood ratio of a positive test = % of patients with the symptom (the test in this instance) who do have migraine divided by the % of patients with the symptom who do not have migraine. Ie, true positive % divided by false positive %
   The specificity of these 3 symptoms is high. Ie, few patients who have these symptoms have HA other than migraine.
3. Despite the origin of the word "migraine" from "hemicrania", unilateral pain is the least sensitive criterion. Throbbing pain is also less sensitive a criterion.
4. The absence of any of the features � nausea, phonophobia, photophobia and exacerbation by physical activity makes the diagnosis of migraine less likely.
5. Migraine precipitants: 1) stress, 2) certain foods, 3) missing a meal, and 4) menses, especially the week before. Food triggers (especially cheese, chocolate) are noted in about half the patients. The positive likelihood ratio of food precipitation = 3.6. Chocolate is the most specific precipitant. However, with the exception of food triggers, the other precipitants are equally prevalent among patients with tension-type HA.
6. Aura: About 1/3 of migraine patients experience an aura before the HA. Visual aura are the most common. Positive visual aura include zigzags, stars, and flashes. Negative aura are scotoma and hemianopsia. Sensory aura occur mainly in the hand and face. (They may progress from hand, to arm, to face). Sensory aura are usually unilateral. Aphasic, and motor aura occur in 5% to 20%. When one of these auras is present, the diagnosis is established. Most aura last less than 30 minutes.
7. Over half the migraine patients have a positive family history. Recurrent vomiting attacks and motion sickness in childhood may predict development of migraine later in life.

1. "The history and results of the general physical and the neurological examinations correctly identify patients who need no further diagnostic evaluation."
2. About 18% of women and 6% of men have migraine HA. In about half, the migraine is severe enough to require bed rest or severely restricted working ability.
3. Mean age of onset of migraine is in the teen years. A first HA occurring after age 40 deserves particular diagnostic diligence to rule out secondary HA.

1. Are the most common HA. Lifetime prevalence is about 75%. Although episodes are less likely to result in loss of work, the total national loss is greater than for migraine because of the greater prevalence of TTHA.
2. Migraine and TTHA can occur in the same patient.
3. The HA is chronic and recurring with few or no migrainous features. They are commonly bandlike, non-throbbing, and bilateral. The great majority last less than 24 hours. (Similar duration to migraine.) Phonophobia and photophobia are uncommon. If either is present, suspect migraine.
4. Except for food triggers, precipitants are similar to those for migraine. (Again if a food trigger is present, suspect migraine.

1. Is distinctive, and easily distinguished from migraine.
2. In contrast to migraine: cluster HA is more common in men. Mean age of onset is age 30, older than for migraine.
3. Pain is commonly periorbital and strictly unilateral. It is usually excruciating. Patients become restless and agitated, and cannot find a comfortable position. Attacks commonly begin during sleep.
4. Autonomic symptoms commonly accompany cluster HA � lacrimation, rhinorrhea, and a partial Horner syndrome.
5. Duration is most commonly 30 to 60 minutes, distinctly shorter than for migraine. (Trigeminal neuralgia pain is much briefer.)
6. Stress and alcohol are common precipitants.
7. Cluster HA derives its name from their periodicity or tendency to cluster over time. Each cluster lasts an average of 1 to 2 months. Patients commonly experience 1 to 2 clusters a year.

Health care workers by definition have an increased exposure to flu.
Protection against an uncomfortable few days of lost-work due to a febrile illness.
Vaccination will increase herd-immunity.

Cost:
My pharmacy quotes a price of $59.00 for ten 75 mg tablets of oseltamivir (Tamiflu).

1. Literature search identified 4 randomized trials of at least 1-year's duration reporting use of drug treatment for patients without known cardiovascular disease. (Primary prevention)
Mean age = 53; initial total cholesterol levels ranged from 220 mg/dL to 290 mg/dL.
Studies used cholestyramine, gemfibrozil, pravastatin (Pravachol) , and lovastatin (Mevacor).
End point = all cause mortality, CHD mortality, and non-fatal myocardial infarction.

1. Drug treatment with cholestyramine and gemfibrozil reduced total cholesterol levels by about 10% ; statin drugs reduced total cholesterol by about 20%.
Drug treatment reduced the odds of a coronary event by 30%.
Drug treatment over 5 to 7 years was not associated with any statistically significant reduction in total all-cause mortality.
Trials using statin drugs suggest a slightly stronger benefit on all outcomes, but did not show a significant reduction in all-cause mortality.

1. In primary prevention trials, lipid controlling drugs reduced the relative risk of CHD events and CHD mortality by about 30% over 5 to 7 years.
The effect on all-cause mortality was small and not significant.
The failure of drug treatment to reduce all-cause mortality in primary prevention trials is most likely due to the generally low risk of mortality in the patient populations that were studied.
"Treatment targeted specifically at primary prevention in patients with higher levels of risk of coronary disease events might reduce all-cause mortality." Lower risk populations might experience a lower all-cause mortality if they were treated for a longer period.
Because the absolute risk of all-cause mortality in primary prevention patients is relatively low (2.4%) over 5 years, the absolute benefit in lives saved will also be low initially. The absolute risk reduction from treatment is proportional to the underlying risk in the person being considered for treatment. 2
The decision about whether to use lipid-controlling drugs for patients without a history of CHD is difficult and requires consideration of outcomes other than all-cause mortality.3
The concomitant use of other drugs (eg, aspirin and beta-blockers), which were not used widely in the trials, may lower absolute risks.

Costs:

For the statin drugs used in this review, my pharmacy quotes:

Pravastatin (Pravachol) 40 mg $356 for 100
Lovastatin (Mevacor) 40 mg $385 for 100.

The trials used 40 mg daily. Thus a month's supply would cost over $100.

1. Randomized trial followed over 400 women with gestational diabetes. All failed to achieve acceptable glucose levels with diet and exercise, and were considered in need of drug treatment.
2. All had singleton pregnancies and a fasting plasma glucose between 95 and 140 mg/dL. Those with levels = 95 were initially treated with diet, but subsequently enrolled in the trial if the glucose rose to over 95.
3. Mean baseline plasma glucose values (mg/dL):
   

   Screen (1-h after 50-g glucose)	169
   Oral glucose tolerance (100-g glucose)
   Fasting	98
   1 hour	199
   2 hour	174
   3 hour	137

4. Randomized at between 11 and 33 weeks to: 1) glyburide, or 2) insulin according to an intensified treatment protocol. Primary end point = desired level of glycemic control.
5. Mean dose of insulin = 85 units/d; mean dose of glyburide = 9 mg/d

1. Mean pretreatment blood glucose concentration was 115 mg/dL.
2. Mean concentration during treatment was 105 mg/dL in both groups.
3. Eight women (4%) in the glyburide group required switch to insulin therapy.
4. No differences in percentage of infants who were large for gestational age; who had macrosomia (birth weight = 4000-g); lung complications; hypoglycemia; admissions to neonatal intensive care; or fetal anomalies.
5. Cord serum insulin concentrations were similar.
6. Glyburide was not detected in cord serum of any infant.

1. Glyburide was as effective as insulin in producing glycemic control.
2. No materno-fetal transfer of glyburide was detected.
3. Past studies with metformin in women with GD reported 18% of infants were large for gestational age, 30% had jaundice; 9% major congenital anomalies.
4. The women in the present study were treated well after organogenesis. Rates of anomalies were similar in both groups and similar to women without diabetes.

1.       Lead-time bias. This pertains to patients who are destined to die at a fixed date regardless of a screen, and any treatment given as a result of the screen. For example, suppose a patient with LC, even in the early stages, is destined to die of the LC on January 11, 2002, regardless of any treatment. (This may be due to metastatic disease already present before resection of the cancer.) A screen discovers the cancer on January 11, 2000. The local cancer is resected. If the screen were not performed, the patient develops symptoms on January 11, 2001. In both cases the patient dies on January 11, 2002. However, those who perform the screen claim that the screen and the resultant surgery prolonged the patient's life by 1 year, and therefore was worthwhile.



2.       Length-time bias. This is a less evident bias. It has to do with the length of time between screens. For example, consider two patients with LC: 1) has aggressive disease, 2) has slowly progressive disease. When an initial screen is done at baseline, both screens are negative. The screen done 2 years later discovers the less aggressive cancer and surgery is performed. However, the patient with more aggressive cancer develops symptoms leading to the diagnosis one year after the original screen (interval cancer). Ie, the cancer is diagnosed, not as a result of the screen, but because of the onset of symptoms. Thus, the aggressive case is excluded from the analysis of the cohort of screened patients. The less aggressive case is included in the screened and treated cohort, and the results of the screening are thus considered to be more favorable than they actually are. Ie, length-time bias leads to elimination of aggressive cancers and favors the discovery of patients with less aggressive disease. This results in a bias favoring screening. Increased detection of more aggressive cancer can be achieved only by shortening the interval between screens, or improving the sensitivity of the test.1
         Contamination of the control population is another common bias in community settings. This occurs when many individuals in the control group actually do not remain untreated controls, but receive treatment outside of the study. Thus, the comparative difference between treated and control group diminishes.
         Overdiagnosis is another consideration. This refers to the fact that some indolent cancers progress so slowly that the patient eventually dies of some other cause. The cancer discovered by screen is not involved. (Over diagnosis is a problem with screening for prostate cancer by PSA.)
         The editorialist goes on to comment that. although CT scanning for LC is more sensitive, the specificity of CT scan for LC is low (many false positives). Thus the predictive value of a positive CT is low. (The predictive value of a positive test is the ratio between true positive tests expressed as a percentage divided by the sum of true positive tests + false positive expressed as a percentage.) This was 12% in one study. This leads to overtreatment with its attendant harms and costs.
         Screening is beneficial only if earlier detection of the disease leads to effective treatment that decreases mortality from the disease or provides longer survival and improved quality of life. Improved survival is a benefit only if it is not iatrogenic pseudosurvival � that is, pseudosurvival because of overdiagnosis, lead-time bias, or length-time bias.
         However, "I believe it should be ethical for physicians to discuss lung cancer screening with selected high-risk patients. As long as the patient understands that screening in no substitute for tobacco avoidance, understands that it is not known whether screening leads to better outcomes, understands that screening may lead to a cascade of other evaluations and potential complications, and is aware that third-party payers may not pay for the tests, it is appropriate for this physician to offer screening if that is what the patient wants."

Since melatonin is considered a dietary supplement, it is not subject to FDA approval and surveillance. I still worry about strength and purity of drugs purchased off the shelf.
The study used "analytic grade melatonin" administered under an Investigational New Drug application.
The study strengthens my belief that melatonin is a legitimate pharmaceutical and has important physiologic actions. RTJ L>

      First, what should physicians not do? Physicians should not prescribe religious beliefs or activities for health reasons. They should not impose their religious beliefs on patients or initiate prayer without the knowledge of the patient's religious background and the likely appreciation of such activity.
      What should physicians do? They should acknowledge and respect the spiritual lives of patients, and should keep interventions patient-centered. Acknowledging the spiritual life often involves taking spiritual history, although this is not appropriate for every patient. For those with illness that threatens life or way of life, it probably is. The American College of Physicians suggests 4 simple questions:

1. Is faith (religion, spirituality) important to you in this illness?
2. Has faith been important to you in the past?
3. Do you have someone to talk to about spiritual matters?
4. Would you like to explore religious matters with someone?

      Taking a spiritual history is often a powerful intervention in itself. Many patients appreciate their physician's sensitivity to these issues. The physician can thus send an important message that he/she is concerned with the whole person, a message that enhances the physician-patient relationship.
      Should physicians pray with patients? It is not appropriate to pray without the patient's consent. Alternatively, prayer can be led by the patient. Our calling as physicians is to "cure sometimes, relieve often, and comfort always". If a distressed patient asks for prayer and the physician sees that such prayer could bring comfort, then it is difficult to justify a refusal to do so.

Practical Pointers can also be accessed on the Internet
The internet version is formatted in HTML. This presents many helpful links, especially when one wishes to use the cumulative index.
The pdf format permits printing out the text in a readable format. RTJ