By J. Mamuk. Pennsylvania State University, Great Valley.
Behavioural change vention refers to both preventing the development of will be best accomplished by influencing the commu- disease as well as its risk factors discount dutasteride 0.5 mg, and primary nity (13) cheap dutasteride 0.5 mg amex. Because of the large number of people near the middle of the distribution, Smoking. Although this is common knowl- North Karelia Project in Finland (14), the Stanford edge, the important role of physicians and other health Three Community Study (15) and ‘‘Live for Life’’ care providers in helping people to stop smoking is less health promotion programme in Sweden (16). The simple The high-risk and population-based strategies are advice from a physician to stop smoking has been far from being mutually exclusive. On the contrary, shown to double the spontaneous rates of quitting in a they are mutually supportive. As many opportunities interventions are more likely to be successful in an as possible in the varied encounters between patients environment where healthy lifestyle habits are widely and the health care system should be used to ask practised. And those who practice high-risk interven- about smoking habits and to offer assistance to those tions are important champions and educators of the who are ready to fight the habit. It should, however, be kept in mind that it is an uphill struggle to give up smoking and there are strong commercial and social forces that promote smoking, especially among the young. In some countries a positive change has been noted with the prevalence of smoking decreasing (5,16). Elevated blood pressure is a well- established preventable risk factor for the development of all manifestations of atherosclerosis, coronary heart disease, stroke, peripheral artery disease and heart Fig. The aim of primary blood pressure is an equally strong risk factor as prevention is to shift the curve toward the left, i. Both mendations emphasise the importance of repetitive sets of guidelines emphasise that recommendations measurements of blood pressure, sometimes over a must be based not only on lipid measurements but also period of several months, for the accurate assessment on assessment of the absolute coronary risk projected of blood pressure Á/ which is needed for the decision by a total risk profile. L may require cholesterol-lowering drug treatment in Fundamental to that decision is the assessment of a patient with high overall coronary risk, whereas a the patient’s overall cardiovascular risk profile, be- serum cholesterol level of 7Á/8 mmol/L may be left cause the detrimental impact of blood pressure is untreated, except for lifestyle advice, in an individual determined by the presence or absence of other risk with low absolute risk (7). Lifestyle calls for lifestyle recommendations for a period of a interventions are important and in some cases can be few months, and if risk reduction is insufficient drug sufficient for adequate control. The safety and tions should be given to all patients considered to have efficacy of statin therapy in the primary prevention hypertension. Weight control, reduction in the use of setting is based on robust trial evidence (31). Drug below 3 mmol/L is ideal for the whole population and treatment is recommended if the systolic blood a worthwhile public health goal is to achieve these pressure is ]/160 and/or diastolic pressure ]/100 levels with appropriate diet and regular physical mmHg, despite lifestyle interventions. Risk factor saturated fats and cholesterol in a given population management in this patient group is extremely im- and the usual levels of serum cholesterol in that portant. Consequently, in viously discussed, a population strategy should be newly published recommendations from the American augmented by the individualised clinical approach of Diabetes Association, the target goal for hypertensive physicians identifying those who need urgent and treatment among diabetes patients is set at B/130/80 aggressive risk factor modification, including drug mmHg, and it is recommended that this should be treatment and family screening. Most recently, the beginning in childhood, should be one of the most Steno-2 study from Denmark (39) has demonstrated important health priorities for the years to come. The intensive treatment relative risk as smoking, hypercholesterolaemia or involved stepwise introduction of lifestyle and phar- hypertension (47). Part of its complex effect may be included reduced intake of dietary fat, regular parti- mediated through enhanced fibrinolytic potential cipation in light or moderate exercise and abstinence and reduced platelet adhesiveness and thus reduced from smoking. Epidemiological stu- even in small doses aspirin can do more harm than dies have shown that the relationship between body good (54Á/56). Although the reduction in relative risk weight and mortality rate is J-shaped, the lowest may be similar in both primary and secondary mortality rate being among those with ‘‘normal’’ prevention, i. Because of this, one should be very careful to Heart Study (44) it was concluded that obesity in use drugs as a general mean of prevention unless the adulthood is associated with a decrease in life ex- benefits have been proven in well-conducted clinical pectancy of about 7 years in both men and women. Increased Several observational studies have suggested that intra-abdominal fat mass, i. Nevertheless, the The recent negative or neutral trial results with lifestyle recommendations given for one risk factor, oestrogen and vitamins (40), in both cases overturning for example hypertension or high serum cholesterol, conclusions from observational studies, demonstrate follow the same general principle as health recom- that in the field of prevention as in therapeutics we are mendations applicable for the public in general, i. The burden of elevated blood pressure, insulin resistance and glucose cardiovascular diseases mortality in Europe. Task Force of the European Society of Cardiology on Cardiovascular intolerance, a prothrombotic state and a proinflam- Mortality and Morbidity Statistics in Europe. Heart and stroke although pharmacological management of insulin statistical update. Dallas, Texas: American Heart Associa- resistance may hold a promise for the future. World another group and also patients met in ordinary Health Stat Q 1988;41:155Á/78. Changes in risk factors explain changes in mortality from Abdominal obesity (waist circumference) ischaemic heart disease in Finland. Singapore and coronary heart disease: a Women /88 cm population laboratory to explore ethnic variations in the Triglycerides /1.
Internal valid- ity is threatened by problems in the way a study is designed or carried out discount 0.5 mg dutasteride visa, or with the instruments used to make the measurements 0.5 mg dutasteride with mastercard. External validity exists when the measurement can be generalized and the results extrapolated to other clinical situations or populations. External validity is threatened when the pop- ulation studied is too restrictive and you cannot apply the results to another and usually larger, population. The results of an internally valid study are true if there is no serious source of bias that can Instruments and measurements: precision and validity 75 produce a fatal ﬂaw and invalidate the study. Truth in the universe relating to all other patients with this problem is only present if the study is externally valid. Improving precision and accuracy In the process of designing a study, the researcher should maximize precision, accuracy, and validity. The methods section detailing the protocol used in the study should enable the reader to determine if enough safeguards have been taken to ensure a valid study. The protocol should be explicit and given in enough detail to be reproduced easily by anyone reading the study. There are four possible error patterns that can occur in the process of measur- ing data. Using exactly reproducible and objective measurements, standardizing the performance of the measurements and intensively training the observers will increase precision. Automated instruments can give more reliable measure- ments, assuming that they are regularly calibrated. The number of trained observers should be kept to a minimum to increase precision, since having more observers increases the likelihood that one will make a serious error. For example, tak- ing a blood pressure is obtrusive while simply observing a patient for an out- come like death or living is usually non-obtrusive. Watching someone work and recording his or her efﬁciency is obtrusive since it could result in a change in behavior, called the Hawthorne effect. If the observer is unaware of the group to which the patient is assigned, there is less risk that the measurement will be 76 Essential Evidence-Based Medicine biased. Blinding creates the climate for consistency and fairness in the measure- ments, and results in reduced systematic error. Non-blinded measurements can lead to differential treatment being given to one of the groups being studied. In single blind- ing, either the researcher or the patient doesn’t know who is in each group. In double blinding, neither the researchers nor subject knows who is in each group. Triple blinding occurs if the patient, person treating the patient, and the researcher measuring the outcome are all blind to the treatment being rendered. Tests of inter- and intra-rater reliability Different observers can obtain different results when they make a measurement. Several observers may measure the temperature of a child using slightly different techniques when using the thermometer like varying the time the thermometer is left in the patient or reading the mercury level in different ways. The researcher should account for variability between observers and between measurements made by the same observer. Variability between two observers or between multiple observations by a single observer can introduce bias into the results. Therefore a subset of all the measurements should be repeated and the variability of the results measured. Inter-observer variability occurs when two or more observers obtain different results when measuring the same phenomenon. Intra-observer variability occurs when the same observer obtains different results when measuring the same phenomenon on two or more occasions. Tests for inter-observer and intra-observer variability should be done before any study is completed. Both the inter-observer and intra-observer reliability are measured by the kappa statistic. The kappa statistic is a quantitative measure of the degree of agreement between measurements. It measures the degree of agreement beyond chance between two observers, called the inter-rater agreement, or between multiple measurements made by a single observer, called the intra-rater agreement. The kappa statistic applies because physicians and researchers often assume that all diagnostic tests are precise. However, many studies have demonstrated that most non-automated tests have a degree of subjectivity in their interpre- tation. It is also present in tests commonly consid- ered to be the gold standard such as the interpretation of tissue samples from autopsy, biopsy, or surgery. Abnormal 10 0 10 100 0 Here is a clinical example of how the kappa statistic applies. He didn’t really feel like reading these and knew that all of his read- ings would be reviewed by the attending. He also reasoned that since this was a screening clinic for young women with an average age of 32, there would be very few positive studies.
Even this may not prevent overly optimistic results from being published buy discount dutasteride 0.5 mg line, and all research must be reviewed in the context of other studies of the same problem buy 0.5 mg dutasteride free shipping. If these other stud- ies are congruent with the results of the study stopped early, it is very likely that the results are valid. Intelligent readers of the medical literature must be able to interpret these results and determine for themselves if they are important enough to ignore in clinical practice. The problem with evaluating negative studies Negative studies are those that conclude that there is no statistically signiﬁcant association between the cause and effect variables or no difference between the two groups being compared. This may occur because there really is no asso- ciation or difference between groups, a true negative result, or it can occur because the study was unable to determine that the association or difference was statistically signiﬁcant. If there really is a difference or association, the latter ﬁnding would be a false negative result and this is a critical problem in medical research. On separate days, each student was given a cup of coffee, one day they got real Java and the next day decaf. After drinking the coffee, they were given a simple test of math problems that had to be completed in a speciﬁed time and each of the students’ scores was then calculated. However, when a statistical test was applied to the results, they were not statistically signiﬁcant, meaning that the results could have occurred by chance greater than 5% of the time. In other words, the researcher concludes that there isn’t a difference, when in fact there is a difference. An example would be concluding there is no relationship between familial hyperlipidemia and the occurrence of coronary artery disease when there truly is a relationship. Another would be concluding that caffeine intake does not increase the math scores of college psychology students when in fact it does. This is a convoluted way of saying that it ﬁnds the alternative hypothesis to be false, when it ain’t! Beta is the probability of the occurrence of this wrong conclusion that an investigator must be willing to accept. Power is the ability to detect a statistically signiﬁcant difference when it actu- ally exists. The researcher can reduce β, and thereby increase the power, by selecting a sufﬁciently large sample size (n). Determining power In statistical terminology, power means that the study will reject the null hypoth- esis when it really is false. By convention one sets up the experiment so that β 132 Essential Evidence-Based Medicine is no greater than 0. Remember, that a microscope with greater power will be able to detect smaller differences between cells. These include the type of variable, statisti- cal test, degree of variability, effect size, and the sample size. The type of variable can be dichotomous, ordinal, or continuous, and for a high power, continuous variables are best. For the statistical test, a one-tailed test has more power than a two-tailed test. The degree of variability is based on the standard deviation, and in general, the smaller the standard deviation, the greater the power. The bigger the better is the basic principle when using the effect size and the sample size to increase a study’s power. These concepts are directly related to the concept of conﬁdence discussed√ in Chapter 10. The conﬁdence formula (conﬁdence = (signal/noise)√ × n) can be written as conﬁdence = (effect size/standard devia- tion) × n. According to this formula, as effect size or sample size increases, con- ﬁdence increases, thus the power increases. Effect of sample size on power Sample size (n) has the most obvious effect on the power of a study with power increasing in proportion to the square root of the sample size. If the sample size is very large, an experiment is more likely to show statistical signiﬁcance even if there is a small effect size. The smaller the sample size, the harder it is to ﬁnd statistical signiﬁcance even if one is look- ing for a large effect size. Remember the two groups of college psychology stu- dents at the start of this chapter. It turns out, when the scores for the two groups were combined, the results were statistically signiﬁcant. For example, one does a study to ﬁnd out if ibuprofen is good for relieving the pain of osteoarthritis. The results were that patients taking ibuprofen had 50% less pain than those taking placebo.
Resistant starch is naturally occurring order dutasteride 0.5mg on line, but can also be produced by the modification of starch during the processing of foods buy dutasteride 0.5mg without prescription. Resistant starch is estimated to be approximately 10 percent (2 to 20 percent) of the amount of starch consumed in the Western diet (Stephen et al. Along the gastrointestinal tract, properties of fiber result in differ- ent physiological effects. Effect on Gastric Emptying and Satiety Consumption of viscous fibers delays gastric emptying (Low, 1990; Roberfroid, 1993) and expands the effective unstirred layer, thus slowing the process of absorption once in the small intestine (Blackburn et al. A slower emptying rate means delayed digestion and absorp- tion of nutrients (Jenkins et al. For example, Stevens and coworkers (1987) showed an 11 percent reduction in energy intake with psyllium gum intake. Postprandial glucose concentration in the blood is thus lower after the consumption of viscous fiber than after consumption of digestible carbohydrate alone (Benini et al. The extended presence of nutrients in the upper small intestine may promote satiety (Sepple and Read, 1989). Fermentation Fibers may be fermented by the colonic microflora to carbon dioxide, methane, hydrogen, and short-chain fatty acids (primarily acetate, propi- onate, and butyrate). Foods rich in hemicelluloses and pectins, such as fruits and vegetables, contain Dietary Fiber that is more completely ferment- able than foods rich in celluloses, such as cereals (Cummings, 1984; Cummings and Englyst, 1987; McBurney and Thompson, 1990). There appears to be no relationship between the level of Dietary Fiber intake and fermentability up to very high levels (Livesey, 1990). Butyrate, a four-carbon, short-chain fatty acid, is the preferred energy source for colon cells (Roediger, 1982), and lack of butyrate production, absorption, or metabo- lism is thought by some to contribute to ulcerative colitis (Roediger, 1980; Roediger et al. Others have suggested that butyrate may be protec- tive against colon cancer (see “Dietary Fiber and the Prevention of Colon Cancer”). However, the relationship between butyrate and colon cancer is controversial and the subject of ongoing investigation (Lupton, 1995). Once absorbed into the colon cells, butyrate can be used as an energy source by colonocytes (Roediger, 1982); acetate and propionate travel through the portal vein to the liver, where propionate is then utilized by the liver. A small proportion of energy from fermented fiber is used for bacterial growth and mainte- nance, and bacteria are excreted in feces, which also contain short-chain fatty acids (Cummings and Branch, 1986). Differences in food composi- tion, patterns of food consumption, the administered dose of fiber, the metabolic status of the individual (e. Because the process of fermentation is anaerobic, less energy is recovered from fiber than the 4 kcal/g that is recovered from carbohy- drate. While it is still unclear as to the energy yield of fibers in humans, current data indicate that the yield is in the range of 1. Physiological Effects of Isolated and Synthetic Fibers This section summarizes the fibers for which there is a sufficient data- base that documents their beneficial physiological human effects, which is the rationale for categorizing them as Functional Fibers. It is important to note that discussions on the potential benefits of what might eventually be classified as Functional Fibers should not be construed as endorsements of those fibers. While plant-based foods are a good source of Dietary Fiber, isolated or synthetic fibers have been developed for their use as food ingredients and because of their beneficial role in human health. In 1988 Health Canada published guidelines for what they considered to be “novel fiber sources” and food products containing them that could be labeled as a source of fiber in addition to those included in their 1985 definition (Health Canada, 1988). The rationale for these guidelines was that there were safety issues unique to novel sources of fiber, and if a product was represented as containing fiber, it should have the beneficial physiological effects associated with dietary fiber that the public expects. The guidelines indicated that both safety and efficacy of the fiber source had to be estab- lished in order for the product to be identified as a source of dietary fiber in Canada, and this had to be done through experiments using humans. Detailed guidelines were later produced for the clinical studies required to assess laxation effects, as this was the physiological function most often used by industry when seeking approval for a novel fiber source (Health Canada, 1997). For each of the fiber sources discussed below, studies will be summarized that relate to one of the three measures of efficacy identified by Health Canada, as these are the three most commonly accepted beneficial effects of fibers. A more complete discussion of these three measures of efficacy may be found later in this chapter. In addition, other potentially efficacious effects will be noted where studies are available. As interest has increased in fiber, manufacturers have isolated various types of fiber from a wide range of carbohydrate sources added to foods. Many of these isolated materials are used as food additives based on func- tional properties such as thickening or fat reduction. As enzymatic and other technologies evolve, many types of polysaccharides will continue to be designed and manufactured using plant and animal synthetic enzymes. Examples in this category include modified cellulose, in which the hydroxyl groups on the glucose residues have been substituted to varying degrees with alkyl groups such as methyl and propyl; fructooligosaccharides manu- factured from sucrose; and polydextrose synthesized from glucose. In some instances, fibers isolated from plants or manufactured chemically or synthetically have demonstrated more powerful beneficial physiological effects than a food source of the fiber polysaccharide. From a meta-analysis of about 100 studies of changes in stool weight with various fiber sources, investigators have calculated the increase in fecal weight due to fiber ingestion (Cummings, 1993). As noted later in this chapter, an increase in fecal weight does not necessarily equate with enhanced laxation, so this needs to be considered in interpreting the results of fecal bulking studies.