7. Clinical Prevention - Anthony F. Jerant

7. Clinical Prevention - Anthony F. Jerant

Background

Definition and Focus

Clinical prevention involves the maintenance and promotion of health and the reduction of risk factors that result in injury and disease. The elements of clinical prevention include screening tests, counseling interventions, and immunizations, as well as chemoprophylaxis, the use of drugs or biologics taken by asymptomatic persons to reduce the risk of developing a disease. This chapter provides the tools for family physicians to meet the formidable challenge of providing clinical prevention services in an evidence-based manner. While mass screening is an important public health tool, the material in this chapter mostly concerns the individualized screening that is offered during single physician-patient encounters. Detailed information regarding lifestyle counseling can be found in Chapter 8. Consistent with the approach of the U.S. Preventive Services Task Force (USPSTF), the focus of this chapter is primary and secondary prevention. Primary prevention is the reduction of risk factors for diseases before they occur, whereas secondary prevention is the identification and treatment of diseases or conditions at an early stage. Both primary and secondary prevention concern asymptomatic individuals. Tertiary prevention, which reduces the future negative health effects of diseases or conditions that have already become symptomatic, is discussed in many other chapters in this book.

The Ongoing Need for Clinical Preventive Services

Tremendous successes in clinical prevention have been realized in the last 50 years. For example, mortality due to coronary heart disease has declined by approximately 50%, and more than half of this decline can be attributed to preventive interventions such as reducing cigarette smoking and detecting and treating hyperlipidemia. The greater than 90% reductions in morbidity and mortality due to measles, mumps, rubella, smallpox, pertussis, tetanus, and Haemophilus influenzae type b resulting from mass vaccination programs are an even greater prevention success story.¹) Nevertheless, the most common underlying causes of death in the United States reflect an ongoing need to improve and expand the delivery of clinical preventive services (Table 7.1).²

The leading health indicators in Healthy People 2010, a blueprint for public health resulting from collaboration between hundreds of state and federal agencies and organizations, were clearly developed with this list in mind (Table 7.2).³ Because the average life expectancy in America lags behind that of nearly 20 other nations, one of the main goals of Healthy People 2010 is to increase life expectancy and years of healthy life. There are 467 specific objectives within 28 focus areas derived from population data, many pertinent to clinical prevention (Table 7.2). The full report is available on the World Wide Web (WWW) at http://www.health.gov/healthypeople . Healthy People 2010 provides a critical link between public health and clinical practice. Approaching every patient with the focus areas in mind will help in detecting the most prevalent contributors to early morbidity and mortality. For example, in the focus area for cancer, one objective is to increase the proportion of adults who receive colorectal cancer screening from 35% to a target of 50%. A 55-year-old patient who has not undergone screening might be informed that while it can reduce the risk of death due to colorectal cancer, only one in three eligible Americans receives such screening. Subsequently, the patient's genetic and environmental history, health habits, and preferences can be used to develop a personalized colorectal cancer prevention plan.

Evidence-Based Clinical Prevention

Principles of Screening

Seven principles should be considered in evaluating a potential screening intervention:

1. The disease or condition in question must lead to substantial morbidity or mortality. Several conditions consistently account for the greatest disease burden in our society. This burden can be quantified using the disability-adjusted life year (DALY), the sum of the years of life lost due to premature mortality and the years of life lost due to disability in a population (Table 7.3).⁴ Screening patients for these conditions and underlying risk factors should be given the highest priority.

2. The screening test employed to detect the condition should be accurate. An ideal screening test has both a high sensitivity (low false-negative rate) and high specificity (low false-positive rate). In practice, screening tests seldom meet this ideal. For example, the CAGE acronym screening tool for alcohol abuse and dependence (see Chapter 59) has fair specificity (76-96%) but relatively low sensitivity (74-78%) at the most commonly used definition of abnormal (two or more affirmative responses). Lowering the abnormal cutoff to one or more affirmative responses would increase the number of problem drinkers detected (sensitivity 86-90%) but would also lead to more false positives (specificity 52-93%).⁵ The trade-off between sensitivity and specificity is a characteristic of all screening tests.

3. The disease or condition should have a high incidence and/or prevalence. Screening is effective for some conditions only when individuals reach a certain age, are of a certain gender, or possess certain risk factors that place them at increased risk for developing the conditions. Stated another way, as the prevalence of a disease or condition increases, the positive predictive value (PPV) of a test increases, regardless of its sensitivity and specificity:

Another way to estimate the overall "yield for effort" of a screening intervention is the number needed to screen (NNS),⁶ which is analogous to the concept of number needed to treat (NNT) in clinical therapeutics. NNS is calculated by taking the reciprocal of the absolute risk reduction (ARR) conferred by screening:

For example, the NNS to prevent one death due to tuberculosis (TB) in programs involving intravenous drug abusers ranges from 103 to 4650, while in studies involving individuals with no identifiable risk factors for TB, the NNS to prevent one death due to TB ranges from 132,690 to 606,797.⁷ These figures provide clinically tangible estimates of the yield of TB screening and reinforce the concept that the PPV of a screening test increases as the incidence of the condition in question increases in the screened population.

4. The disease or condition should have an asymptomatic period during which it can be detected. Diseases with long asymptomatic periods, such as cervical cancer, are easier to target with screening than diseases with a short preclinical duration, such as leukemia. However, for conditions with a long preclinical duration, lead-time bias can make it appear that a group of screened patients survives longer than a group that is not screened. In reality the screened patients may simply be finding out they have the disease earlier, during its asymptomatic phase. To avoid attributing benefit to a screening program that suffers from lead-time bias, screening decisions should be based on comparisons of actual mortality rates, rather than on measures that are affected by the time elapsed since diagnosis, such as 5-year survival rates. A second problem related to preclinical disease duration is called length-time bias. Less aggressive cases have a longer asymptomatic period and are more likely to be detected by screening than more aggressive cases. Thus, a screening program may appear to improve survival when it is actually only detecting more indolent cases that have a better prognosis. Prostate cancer screening has been criticized for many reasons, including strong concerns about lead-time and length-time bias.

5. The disease or condition should have a widely available and acceptable treatment known to improve outcomes. Many conditions that are otherwise worthy candidates for screening are not currently amenable to treatments that change their natural history. For example, dementia accounts for a substantial number of DALYs (Table 7.3) but fails to meet this criterion.

6. The screening procedure should entail reasonable health risks and financial cost. Screening cost estimates should include not only the cost of an initial screening test but also costs related to repeat office visits, specialty referrals, additional testing, false positives, and complications. Formal cost-effectiveness analyses of preventive interventions account for all of these factors. The end point of such analyses is often the ratio of dollar cost per quality-adjusted life year (QALY), the product of the number of years of life and the quality of those years as measured from 0 (indifference between life and death) to 1 (full health) on a questionnaire. Thus, a screening test that provides an average of 12 more years of life with a quality rating or utility of 0.4 is said to provide 4.8 QALYs. Although there is no universal agreement on what dollar cost/QALY ratio cut-point defines a cost-effective screening program, Table 7.4 provides a comparative listing of ratios for some widely accepted preventive practices.⁸ Cost-effectiveness must be considered from the societal perspective, but clinicians can greatly influence the costs of screening programs by taking an evidence-based approach.

7. The screening procedure should be acceptable to the patient and society. The yield of a screening program is decreased if many candidates are unwilling to undergo testing. For example, colorectal cancer screening via flexible sigmoidoscopy is supported by research evidence, but the rate of patient adherence to a physician's recommendation for flexible sigmoidoscopy is only 35%, partially due to test discomfort and inconvenience.

Family physicians can effectively individualize the following general clinical prevention guidelines by considering each of these seven screening principles and the way they apply to their specific practice settings and patient populations.

Clinical Preventive Services Guidelines

The 1996 recommendations of the USPSTF, found in the Guide to Clinical Preventive Services, 2nd edition,⁹ were chosen as the primary resource for this section for several reasons. First, the recommendations are generated using an explicit evidence-based approach, and the items listed in the age-specific recommendation tables are those for which the USPSTF concluded that there is either good or fair evidence to support the recommendation. Second, in contrast to the recommendations of organizations such as the American Cancer Society (ACS) and medical professional groups, the recommendations are not directly tied to public awareness efforts and professional or political agendas. Finally, the USPSTF makes recommendations throughout the life cycle and is thus highly relevant to family physicians. A revised third edition of the guide is scheduled to appear in 2002, and new recommendations and updates are being posted on the WWW as they are released; go to www.ahrq.gov/clinic/uspstfix.htm .

General Recommendations for All Age Groups

In general, new patients should undergo a comprehensive history and physical examination, have a health risk appraisal completed, and be educated regarding age-specific preventive services. Previous health records should be obtained to avoid duplication of services, and additional services may be added routinely based on the individual's risk profile. Because the evidence base is continually growing and changing, family physicians must frequently update their clinical prevention protocols as new evidence becomes available.

Birth to Ten Years (Table 7.5)

Period Immediately Following Birth

Screening for congenital conditions is the first priority in prevention for newborns. All 50 states require testing for phenylketonuria and congenital hypothyroidism, but states vary regarding other mandated tests. In addition to mandated screening tests, infants born to mothers at risk for human immunodeficiency virus (HIV) infection but whose infection status is unknown should be considered for HIV testing. From the time of birth and throughout childhood, it is important to be aware of family psychosocial and socioeconomic factors such as poverty and parental substance abuse that place children at increased risk for multiple adverse health outcomes and developmental problems. For example, parents who smoke tobacco must be counseled regarding the risks to infants of passive smoke exposure, including higher rates of otitis media and lower respiratory tract infections. Counseling of smoking mothers has been shown to reduce their children's exposure to environmental tobacco smoke, regardless of the mothers' eventual cessation status.¹⁰

Passive tobacco smoke exposure is also associated with an increased risk for sudden infant death syndrome (SIDS),¹¹ a leading cause of death in this age group. The USPSTF has not produced a recommendation regarding optimal infant sleep position, but substantial evidence suggests that SIDS is associated with the prone sleep position. Further, although no definite causal link has been established, populations in which physician counseling and media efforts has led to increased use of the supine sleeping position have observed decreased rates of SIDS, resulting in an American Academy of Pediatrics (AAP) recommendation that physicians counsel all parents to place infants to sleep on their backs on a firm surface.¹² A dialogue regarding breast-feeding should ideally be begun during the early prenatal period. Nevertheless, because it is associated with lower rates of otitis media and infectious diarrhea,¹³ physicians should encourage all mothers to breast-feed at the time of birth. This protective effect follows a dose-response relationship so that infants who are not exclusively breast-fed still benefit. Newborns should ideally receive their first hepatitis B vaccination, using a thimerosal-free formulation, prior to discharge from the hospital.

Infancy to Age Two

Ensuring that appropriate growth is being maintained is an important preventive task in this group. Very low birth weight children, an increasing population, often have postnatal growth rates that lag behind those of term infants. Special growth curves, produced by several formula manufacturers, should be utilized until "catch up" growth is achieved, usually at about 3 years of age. Injury prevention counseling should also be emphasized. Injuries account for two of every five deaths in children aged 1 through 4, four times the number of deaths due to birth defects, the second leading cause of death in this age group. Clinicians must also remain alert to the various presentations of family violence, which may include injuries initially attributed to accidents. While the debate regarding lead screening in childhood continues, the USPSTF and Centers for Disease Control and Prevention (CDC) currently recommend a selective approach. Children should be screened if they live in areas with risk for lead exposure, belong to groups that may be at risk (such as the poor), or are found to be at risk based on a "yes" answer to any of the following three questions: (1) Does the child live in or regularly visit a house that was built before 1950? (2) Does the child live in or regularly visit a house that was built before 1978 with recent (within the last 6 months) or ongoing renovations or remodeling? (3) Does the child have a sibling or playmate who has or did have lead poisoning? Physicians must also be aware of local policies, since some states mandate screening.

Table 7.6 provides the most recent universal childhood immunization schedule.¹⁴ Recent changes include the addition of pneumococcal conjugate vaccination and, in certain areas, hepatitis A vaccination. Unfortunately, many children receive immunizations late or not at all, placing them at risk for infectious diseases and increasing the chance of community infectious disease outbreaks in vaccinated individuals.¹⁵ Physician failure to review immunization status at each visit and unnecessary practice policies against vaccination in certain circumstances, such as in the presence of acute minor illness with low-grade fever, are important causes of missed opportunities to vaccinate.¹⁶ Evidence-based vaccination protocols, provider education, and immunization flow sheets may help to reduce missed opportunities. A "catch-up" schedule should be employed for children who have fallen behind to rapidly return them to full coverage (Table 7.7).¹⁷ The Immunization Action Coalition produces excellent resources for both physicians and parents on the WWW: http://www.immunize.org/ .

To help prevent dental caries, children who live in communities with low levels of fluoride in the water should be prescribed fluoride supplements beginning at 6 months of age. Other dental preventive efforts include counseling parents to put children to bed without a bottle and recommending periodic dentist visits beginning at around age 3.

Two to Ten Years

Early detection of cardiovascular disease risk factors should be a major focus of screening beginning in early childhood. The body mass index (BMI) is a practical indicator of the appropriateness of weight for height in children age 2 and older and can be plotted on recently updated growth curves. Although a low BMI can indicate poor nutrition or an underlying medical disorder, elevated BMI in childhood is a more common problem that is reaching epidemic proportions in the U.S. For children 6 and older, a BMI from the 85th to the 95th percentile indicates overweight, whereas a BMI above the 95th percentile indicates obesity. Childhood obesity is associated with a host of immediate and long-term health risks, including increased rates of obesity and early mortality in adulthood.¹⁸ Early identification should be followed by frequent monitoring and parental counseling regarding appropriate diet and nutrition (also see Chapter 53). Physicians should also screen children for a sedentary lifestyle, a major contributor to childhood obesity, and provide counseling regarding physical activity. All children should receive periodic blood pressure measurement throughout this period, and those with measurements that persistently exceed the 95th percentile values in tables based on gender, age, and height should receive further evaluation. Such tables are available on the WWW: http://www.nhlbi.nih.gov.ezp-prod1.hul.harvard.edu/health/prof/heart/hbp/hbp_ped.htm .¹⁹The USPSTF and other organizations recommend cholesterol measurement only in children at high risk for adult coronary artery disease. Risk factors include a family history of premature cardiovascular disease or family members with cholesterol levels greater than 240 mg/dL.

Injury prevention counseling should be continued throughout this period. Thirty-three percent of injuries in this age group are due to violence, and 67% are due to unintentional injuries. Simple measures that reduce injury-related mortality in children, such as the use of helmets when bicycling, should be emphasized. Firearm safety should also be reviewed. Safe sun precautions should be periodically reviewed for children at increased risk for skin cancer, including those with a family history, a large number of moles, atypical moles, poor tanning ability, or light skin, hair, and eye color. The immunization series outlined in Table 7.6 should be continued as appropriate throughout childhood, so that all children will have received the full complement of vaccinations by the age of 12. Because purified protein derivative (PPD) testing of all children is exceedingly expensive and results in many false-positive tests, the USPSTF and CDC recommend a selective approach to screening based on the risk of exposure to TB. Exposure risk factors include birth or prior residence in a region where TB is highly prevalent, such as Southeast Asia, and close exposure to persons known or suspected to have TB.²⁰ (also see Chapter 84).

Table 7.6. Recommended Childhood Immunization Schedule United States, 2002

Eleven to Twenty-Four Years (Table 7.8)

This period includes adolescence, a developmental period that poses unique clinical prevention challenges (also see Chapter 22). Although comprehensive guidelines for preventive care in this age group have been proposed, evidence to support many of the items included is lacking. It is especially unclear whether physician counseling is capable of changing adolescent health behaviors and impacting on key adverse health outcomes. An important principle of prevention for this age group is opportunistic delivery of services. Since adolescents seldom visit a physician specifically for preventive care, every clinic visit by an adolescent should be viewed as an opportunity to provide prevention. Unfortunately, very low rates of clinical preventive services delivery have recently been observed for the typical adolescent visit.²¹ Although adolescents may initially be hesitant to discuss health risk behaviors, they appear to become more willing to do so with repeated physician efforts.²²Appointment invitation letters can increase the number of visits made by adolescents specifically to receive preventive services.²³

Between 50% and 75% of all deaths in this age group are due to unintentional injuries, suicides, and homicides. Providing brief counseling regarding proven injury prevention measures is prudent. Important recommendations regarding motor vehicle injury reduction might include not driving at night for the first year after a driver's license is obtained, not riding in a car with an intoxicated individual, and always using a three-point seat restraint.²⁴ Cardiovascular risk reduction measures such as recommending tobacco avoidance and regular exercise and screening for obesity should be continued. For sexually active teens, contraception and sexually transmitted disease (STD) avoidance counseling are critical. The third USPSTF has released an advance statement regarding chlamydia, the most common STD in the United States, recommending screening all women who are sexually active and aged 25 or younger; have more than one sexual partner; have had an STD in the past; and do not use condoms consistently and correctly, regardless of age. Periodic screening for other STDs in sexually active teens and young adults should also be considered.

Because most alcohol problems begin in early adulthood, the USPSTF recommends screening for problem drinking for all adolescents and young adults using either "careful historytaking" or a standardized questionnaire such as the CAGE. Although finding insufficient evidence to recommend for or against routine screening for other drug abuse, given the increasing prevalence of amphetamine and other illicit drug use in many areas, physicians should have a low threshold for questioning young people about drug use (also see Chapter 60).

In addition to ensuring that a tetanus booster is administered at about 10 years after the last childhood tetanus vaccination, physicians should inform college students about the increased risk of meningococcal infection in crowded dormitory settings and provide them with information regarding meningococcal vaccination.²⁵

Twenty-Five to Sixty-Four Years (Table 7.9)

Women's Health Issues

Although preconception counseling is important for all young women, an opportunistic approach must be taken since few specifically request such care (see Chapter 10). Women planning pregnancy or at risk for unintended pregnancy should be advised to take folic acid, 0.4 to 0.8 mg/day, beginning at least 1 month prior to conception and continuing throughout the first trimester of pregnancy to reduce the risk of neural tube defects. This dose can be obtained by taking a prenatal vitamin daily. Physician advise about folic acid has been shown to dramatically increase patient compliance with this recommendation.²⁶ Screening for cervical cancer using the Papanicolaou (Pap) smear is recommended every 1 to 3 years for all women who have been sexually active and who have a cervix. Although the most cost-effective interval for repeat testing is controversial, the most important things physicians can do to reduce the incidence of cervical cancer are to ensure that as many women as possible receive at least some screening and to ensure that abnormal results are followed up appropriately. Of women who develop invasive cervical cancer, 50% have never had a Pap smear, 10% have not had a Pap smear within 5 years of diagnosis, and 10% have not received appropriate follow-up of a prior precancerous result.²⁷

Breast cancer screening should be offered as women enter middle age, but the optimal time of initiation remains an emotionally charged, controversial issue (see Chapter 107). The USPSTF recommends screening with mammography alone or mammography plus clinical breast examination (CBE) for all women of ages 50 to 69. The task force found insufficient evidence to recommend for or against mammography or CBE for women of ages 40 to 49 or 70 and older, and for teaching patient breast self-examination at any age. In 1997, a National Institutes of Health (NIH) consensus panel initially issued a statement agreeing with the USPSTF position. Shortly after, following a storm of rebuttals by academicians, politicians, and professional interest and advocacy groups, the panel reversed its statement, recommending initiation of periodic mammography and CBE for all women beginning at age 40, as is advocated by the ACS. Unfortunately, these conflicting recommendations and the complexity of the medical literature in this area have greatly confused patients and physicians. It is clear that the potential mortality benefit from breast cancer screening in women of ages 40 to 49 is much smaller than that obtained by screening women of ages 50 to 75, and that beginning screening at an earlier age results in a higher lifetime incidence of false-positive tests.²⁸For now, physicians must review the evidence, form their own conclusions, and then use an "informed consent" approach in negotiating a plan with patients. The National Cancer Institute's Breast Cancer Risk Assessment Tool may help in developing individualized recommendations: http://bcra.nci.nih.gov.ezp-prod1.hul.harvard.edu/brc/ .

Physicians should also provide counseling to reduce the risk of osteoporosis by encouraging women to remain physically active, consume 1000 to 1500 mg of calcium daily, and avoid tobacco use. Bone density measurement may be indicated in women with significant risk factors for osteoporosis such as Caucasian ancestry, petite body frame, low body weight, tobacco use, excessive alcohol and caffeine intake, and prolonged corticosteroid use (see Chapter 122). During the perimenopause, discussion regarding hormone replacement therapy (HRT) should be initiated. Although long-term HRT reduces the risk of osteoporosis and associated fractures, its use is associated with a slight increase in the incidence of breast cancer, and its potential benefit in the primary and secondary prevention of cardiovascular disease remains unproved. Thus, an "informed consent" approach to counseling is advised, with careful weighing of patient preferences and risk factors for osteoporosis, heart disease, and breast cancer.

Men's Health Issues

Prostate cancer is the second leading cause of death for men over age 55, and African-American men have a slightly higher incidence of prostate cancer than other men. While acknowledging its clinical importance, the USPSTF found a lack of evidence to recommend for or against screening for prostate cancer with digital rectal examination (DRE), serum prostate-specific antigen (PSA), or other tests. Evidence that early diagnosis of prostate cancer improves long-term survival is lacking, and there are potential costs and psychological burdens related to the expected high number of false-positive screening tests is large²⁹ (see Chapter 98). Refinements in PSA testing are promising but have not yet been properly evaluated. Despite these concerns, the ACS recommends annual DRE for all men starting at age 40, annual PSA testing beginning at age 40 for African-American men and those with a history of prostate cancer, and annual PSA testing beginning at age 50 for all others. The lack of a clear evidence base for prostate cancer screening and the conflicting recommendations of various organizations have created confusion among physicians and patients alike. As with breast cancer screening in women under age 50, an "informed consent" approach to counseling and educating patients should be utilized. Because prostate neoplasms usually grow slowly, men with a life expectancy of less than 10 years should generally not be screened.

Issues of Importance to Both Men and Women

In addition to its importance as a major cardiovascular disease risk factor, tobacco use has been linked to increased risk for cervical, bladder, lung, and other cancers. Strong counseling regarding smoking cessation, adequate physical activity, and a prudent diet are part of general cancer prevention efforts (see Chapter 8). The USPSTF found insufficient evidence to recommend for or against routine screening for skin cancer by primary care providers or counseling patients to perform periodic skin self-examinations. However, because one in six Americans will develop skin cancer during their lifetime and the incidence of malignant melanoma has increased rapidly during the past decade, physicians should briefly assess skin cancer risk in all individuals (see Chapter 117). Those at increased risk should be advised to avoidance of sun exposure, particularly between 10 A.M. and 3 P.M., and to use protective clothing such as shirts and hats when outdoors. The USPSTF found insufficient evidence to recommend for or against advising sunscreen use. For patients at increased risk for malignant melanoma, such as those with familial atypical mole and melanoma syndrome, referral to a skin cancer specialist for evaluation and surveillance should be considered.

Screening for colorectal cancer should be offered to all average-risk men and women beginning at age 50 (see Chapter 92). The USPSTF recommends annual fecal occult blood testing (FOBT), periodic flexible sigmoidoscopy (FS), or both, stating that there is insufficient evidence to make more specific recommendations. Colonoscopy can detect proximal adenomas and neoplasms, but it is more expensive than FS, is associated with a higher risk of complications such as perforation, and has not been shown to be superior in reducing colorectal cancer mortality. Modeling studies suggest that annual FOBT combined with FS every 5 years, beginning at age 50, is the most cost-effective approach to screening and may reduce colorectal cancer mortality by 50% to 80%.³⁰ As for cervical cancer screening, the major focus in colorectal cancer detection should be to ensure that as many eligible people as possible receive at least some type of screening. Less than half of eligible patients have undergone FOBT or FS within the preceding 5 years.³¹ Medicare provides reimbursement for screening FOBT and FS and, beginning in July 2001, will also reimburse for screening colonoscopy once every 10 years. Even a single colonoscopy at 55 years of age may reduce colorectal cancer mortality by 30% to 50%.³⁰ Patients who are reluctant to undergo colorectal cancer screening may be willing to have "once in a lifetime" screening. More aggressive screening should be considered for those at increased risk for colorectal cancer, such as those with a family history of colorectal cancer or adenomatous polyps.³²

Outside of cancer screening measures, cardiovascular disease prevention should be the major focus of preventive efforts in this age group, including periodic blood pressure screening. The third USPSTF has issued an advanced recommendation to periodically test total cholesterol levels in all men of ages 35 and older and all women of ages 45 and older (see Chapter 119). This extends the recommendations of the second USPSTF, which supported routine cholesterol screening only through age 65. High-density lipoprotein (HDL) and low-density lipoprotein (LDL) screening is recommended for individuals at high risk for cardiovascular disease. The American College of Physicians (ACP) recommends periodic total cholesterol screening in men of ages 35 to 65 and women of ages 45 to 65, with follow-up HDL testing for individuals with elevated levels.³³ Treatment decisions in the ACP recommendations are based on the ratio of total to HDL cholesterol, based on research indicating that higher ratios confer increased risk for cardiovascular disease. By contrast, the National Cholesterol Education Program's (NCEP) Adult Treatment Panel III recommends that a routine fasting lipoprotein profile (total, HDL, and LDL cholesterol and triglyceride levels) be obtained every 5 years in all adults of ages 20 or older.³⁴ As for the cancer screening controversies outlined above, physicians must weigh the evidence supporting each recommendation and collaborate with patients to determine the appropriate course of action.

Tobacco cessation counseling should be provided when applicable, and information on a low-fat diet that is rich in fresh fruits and vegetables and on regular physical activity should be conveyed. The incidence of obesity is increasing at an alarming rate in the United States, conferring increased risk for major cardiovascular risk factors such as hypertension and elevated cholesterol. Periodic weight and height assessment and BMI surveillance should be provided, with further evaluation and intervention offered to those individuals who are overweight (BMI 25.0-29.9) or obese (BMI ≥30.0). Although the USPSTF found insufficient evidence to recommend for or against screening for diabetes mellitus in asymptomatic adults, given its association with obesity and its role as a cardiovascular risk factor, physicians should have a low threshold for obtaining screening fasting serum glucose levels (see Chapter 120). Diabetes screening should also be considered for those with a family history of diabetes and those from high-risk ethnic groups, including Hispanics and Native Americans. Physicians often have difficulty determining the overall level of cardiovascular disease risk for individuals of varying age and either gender in the face of multiple risk factors. Coronary disease risk prediction score sheets which account for multiple variables, may be useful in this regard.³⁵ The score sheets are also available on the WWW: http://www.nhlbi.nih.gov.ezp-prod1.hul.harvard.edu/about/framingham/riskabs.thm . Derived from the predominantly white, middle-class Framingham Heart Study population, they may be less accurate when applied to other types of individuals.

The USPSTF found insufficient evidence to recommend for or against routine aspirin prophylaxis for the primary prevention of myocardial infarction or stroke. Because it is associated with a small increase in the risk of hemorrhagic stroke,³⁶ aspirin chemoprophylaxis should be employed mostly for those patients with risk factors for cardiovascular disease. Moderate alcohol consumption may reduce the risk of cardiovascular disease, but routine physician endorsement of moderate alcohol use for patients who are not already drinking is not recommended given the high prevalence of problem drinking in the U.S. Indeed, the USPSTF recommends screening for problem drinking in all adults and questioning regarding other drug abuse in those considered at increased risk.

Age 65 and Older (Table 7.10)

This group includes both the "young old" (ages 65 to 79) as well as the "oldest old" (ages 80 and beyond), which is now the fastest growing segment of the U.S. population (also see Chapters 23 and 24). However, there is tremendous physiologic variability in the elderly that makes recommendations for prevention based on age alone risky. In both chronologic and physiologic terms, aging impacts on some of the criteria for preventive interventions outlined earlier in this chapter. For example, prostate cancer screening is not indicated for many individuals in this group given the long interval between detection via screening and the earliest time of expected impact on mortality. In addition, older adults may wish to focus primarily on quality of life during their remaining days. Screening interventions that are associated with inconvenience and discomfort may not be desired, regardless of their potential to reduce mortality. Finally, there is a limited evidence base to support many preventive interventions in this age. Physicians must discuss these gaps in evidence, the risks and benefits of screening, and the quality of life goals of older adults before embarking on screening interventions.

The USPSTF recommends annual influenza vaccination as well as a single immunization against Streptococcus pneumoniae for all adults of ages 65 and older. Periodic vision and hearing screening are also suggested because the incidence of both functional vision and hearing problems increases dramatically with aging, rising from about 10% at age 65 to approximately 40% by age 90. Injuries, particularly falls, remain an important source of morbidity and mortality in this group but are more likely to occur while performing simple daily tasks such as walking to the bathroom at night. Fall prevention measures including regular exercise, environmental hazard reduction, and avoiding sedating medications should be discussed with all older individuals. Those who are frail, have had prior falls, or are at ongoing high risk for falls may benefit from a multifactorial intervention that includes home assessment and a hip-protective undergarment.³⁷ End of life planning is also an important preventive care topic for older patients. The value of medical advance directives in improving primary care physicians' and lay surrogates' accuracy in predicting a patient's wishes for care is unclear. However, advance directive discussions and documentation may improve the prediction of patients' wishes by hospital-based physicians and may improve patients' sense of well-being and satisfaction with care.³⁸ Finally, many elders live in poverty, and many reside in assisted living and skilled nursing facilities. These older adults are often frail and may face substantial socioeconomic disadvantages. Screening these individuals for nutritional adequacy, social isolation, depression, and the ability to perform basic and instrumental activities of daily living should be considered.

The Process of Delivering Preventive Care

The Move Toward Accountability in Preventive Services Delivery

Physicians are now being held accountable for offering and delivering evidence-based preventive services. Quality of care models such as the Health Plan Employer Data and Information Set (HEDIS) seek to provide health care purchasers and consumers with a standard against which individual plans can be compared and evaluated. The HEDIS 2001 measures are heavily weighted toward clinical prevention, including items such as breast cancer screening rates, childhood immunization status, and rates of advising smokers to quit. Health plans and clinicians that fail to meet quality thresholds for these indicators are at risk for declining patient enrollment as consumers transfer their care to "higher performers." Nevertheless, delivering individualized, evidence-based clinical preventive services remains a formidable challenge. This section provides a list of issues hindering the delivery of optimal clinical preventive services and provides potential solutions suggested by the research literature.

Organizational Issues and Potential Solutions

Issue 1: Time Constraints of the Clinical Encounter

There is a finite amount of time that can be spent with each patient, and in this time the physicians must address a range of concerns in addition to providing clinical preventive services. In the landmark Direct Observation of Primary Care (DOPC) study, one third of 4401 patient encounters included discussion of at least one preventive service, but only 3% of the all encounter time was allotted to preventive services.³⁹ Time pressures will increase with the aging of the population, as more patients present with multiple chronic diseases, conditions, and functional limitations.

Potential Solutions. Physicians must employ the incremental approach to clinical prevention that is endorsed by the USPSTF. The most urgent priorities for prevention can be addressed first, leaving others for future encounters. Standard "scripts" or minipresentations concerning common preventive topics may increase efficiency.

Issue 2: Limited Dissemination of New Findings and Evidence-Based Prevention Guidelines

The dissemination of evidence-based prevention guidelines in textbooks and journals has limited impact. Such resources, while valuable, rapidly become out of date and may not be readily available at the point of patient care.

Potential Solutions. Evidence-based summary resources that present up-to-date information in a rapid-use format include Patient-oriented evidence that matters (POEMs), http://www.jfponline.com; clinical evidence, http://www.clinicalevidenceonline.com/; the Cochrane Library, http://www.updateusa.com/clibhome/clib.htm; and the ACP Journal Club, http://www.acponline.org.ezp-prod1.hul.harvard.edu/journals/acpjc/jcmenu.htm. The Internet is already an established tool for the delivery of recommendations at the point of care. In the near future, palm-top computers will allow even better point of care access to recommendations.

Issue 3: Competing and Conflicting Recommendations

Many organizations publish recommendations advocating clinical preventive services that are not evidence-based. Clinicians may become confused by conflicting guidelines and are often faced with patients requesting interventions that are promoted by these organizations but not supported by rigorous evidence.

Potential Solutions. The USPSTF recommendations should be utilized whenever possible, and patients should be informed about the levels of evidence for specific interventions. Prevention plans that account for local practice characteristics and patient risk factors, preferences, and beliefs can then be negotiated.

Issue 4: Lack of Office Systems Organized to Provide Effective Preventive Services

Office systems used by practices with successful prevention efforts include designated roles for staff at all levels, paper and computer-based health risk appraisal tools, reminder systems, patient education materials, and record systems, and a quality monitoring and improvement process.

Potential Solutions. The best-known set of materials aimed at improving clinical prevention is Put Prevention into Practice (PPIP). The PPIP kit is paper-based and includes flow sheets, patient-held prevention records, a clinician handbook, prevention prescription pads, medical record reminder stickers, patient reminder postcards, and posters for waiting and examination rooms. Implementing the PPIP office system has been shown to modestly increase the rates of delivery of multiple USPSTF-recommended preventive services. However, dissemination of PPIP has been slow and limited, the absolute increase in rates of delivery for specific services is small, and the positive effects related to its implementation diminish beyond 1 year of follow-up.⁴⁰ Both paper and computer-based reminder systems, including those linked to comprehensive electronic medical records, have been shown to improve rates of preventive services delivery, and the impact appears greatest when the reminder is provided to the physician at the time of a patient visit.⁴¹ As for PPIP, the number of practices utilizing such resources is small and their absolute impact has been limited.

The smaller than anticipated impact of these tools has led to the recognition that the problem of low preventive service delivery rates is a complex, systems issue. The DOPC study suggests there are two major differences between practices delivering limited preventive services and those providing higher levels of these services: (1) the degree of pro-activity in dealing with competing practice demands, and (2) physician philosophy.⁴² Practices with the greatest need to improve preventive care may be the least likely to implement programs like PPIP due to overwhelming competing demands, such as a practice that is heavily weighted toward acute medical care or physicians with a low "prevention orientation." Developing and testing approaches to dealing with competing demands in primary care and changing physician behavior should be given the highest priority. In the meantime, adapting generic materials to individual practice circumstances and enlisting nonphysician clinic staff in prevention efforts are useful first steps. For example, modification of PPIP flow sheets to meet local needs may result in better acceptance of the materials and higher rates of flow sheet completion,⁴³ and simple mailed or telephone call reminders provided by nonphysician staff can increase childhood immunization "up to date" rates.⁴⁴ In the future, remote home-based health risk appraisals, conducted using the Internet and other distance communications technologies, are likely to become routine.

Issue 5: Poor Reimbursement for Preventive Services

In 1988 less than 5% of health care expenditures in the United States was allocated to prevention, and only one third of those expenditures were allocated to clinical prevention.⁴⁵ Tobacco cessation counseling and hearing, vision, and blood pressure screening are all endorsed by the USPSTF for older adults, yet none are covered by Medicare. Paradoxically, many states mandate coverage for screening services not recommended by the USPSTF.

Potential Solutions. Physicians must remain advocates for a preventive health care agenda, making sure local congressional representatives and health plans are aware of shortfalls and misplaced priorities.

Physician and Patient Issues and Potential Solutions

Issue 1: Failure to Adopt and Maintain a Prevention Orientation

Despite the proven benefits of many clinical preventive efforts, some physicians have a practice style that de-emphasizes prevention. In the DOPC study, physicians with a higher volume practice had lower up-to-date rates of preventive screening and counseling services and immunizations.⁴⁶ Female physicians have consistently been shown to offer more clinical preventive services than male physicians, and the effect is not limited to gender-specific interventions.⁴⁷ In addition, some patients do not embrace the concept of clinical prevention.

Potential Solutions. All physicians, and particularly males and those working in high-volume settings, should carefully examine their practice style to ensure it is prevention-centered. Physicians must open a dialogue with patients who do not have a prevention orientation by providing individually tailored information and collaborating to determine the areas in which the patient is most ready to accept preventive interventions.

Issue 2: Holding on to Non-Evidence-Based Beliefs and Practices

Given the time constraints of the modern clinical encounter, it is critical to discard disproved and questionable preventive practices. Focusing on such services reduces the amount of time and money that can be devoted to providing evidence-based services and compounds many of the issues listed above. For example, the "complete physical" appointment accounts for as much as one third of physicians' time spent seeing patients in some practices, yet many elements of this venerable activity have no proven value.

Potential Solutions. Physicians must let go of non-evidence-based prevention ideas as part of the solution to the competing demands issue. Since many patients never make check-up visits, preventive services are best delivered over time, during acute illness and other visits. Making the shift away from the "complete physical" model will require patient education, since people have come to expect certain low-yield maneuvers and interventions. A caring, "high-touch" manner can be conveyed to patients without resorting to the misleading reassurance of a normal heart and lung examination. Patients who request non-evidence-based interventions should be congratulated for their interest in prevention and their health. The dialogue should focus on the reasons for the patient's concern about the health issue in question. The evidence to support the intervention should be summarized and placed in the context of the individual. Finally, a prevention plan is negotiated. Although some patients may still insist on non-evidence-based interventions, most will be satisfied with this approach.

Issue 3: Failure to Account for Varying Patient Health-Belief Models

The United States is increasingly multicultural, and culture and ethnicity impact on every aspect of preventive care, from genetics to health behavior. Some traditional cultural health belief models do not include the Western construct of the concept of prevention.

Potential Solutions. Physicians should learn about the ethnic groups, cultures, and socioeconomic strata represented in their patient population. A rapid overview can be obtained using the U.S. Census Bureau's WWW site at http://www.census.gov, which includes color maps and tables detailing the ethnic distribution of local neighborhoods, language spoken at home, and aggregate family incomes. Becoming involved in community cultural and ethnic activities is an important next step. Perhaps the most important skill in providing multicultural care is to approach each patient without relying on cultural stereotypes. Differing degrees of acculturation and interindividual variability in beliefs make such generalizations dangerous.

Issue 4: Poor Preventive Communication Skills

Just as physicians must learn key physical examination and history-taking skills to diagnose acute medical illnesses, they must also acquire and maintain the communication skills needed to provide optimal clinical prevention. These skills include the ability to (1) translate research and statistics into lay terms, (2) determine patient readiness to modify a health risk behavior, and (3) negotiate a clinical prevention plan.

Potential Solutions. Health systems increasingly offer communication skills training to physicians, recognizing that deficiencies result in poorer health care outcomes and higher costs. Although the best method of conveying health risk information to patients remains unclear, helping patients to understand how a health problem develops (its antecedents) and to recognize what could happen to them as a result (its consequences) may be more successful than simply providing numerical risk information.⁴⁸ In determining a patient's readiness to change a risk behavior, the transtheoretical model provides a useful framework (Table 7.11).⁴⁹ The model illustrates that changes in behavior occur gradually, through a predictable series of steps. Individuals seldom skip steps, so that the physician's role is to assist them in moving to the next stage of change rather than to push them toward behavior change in one giant leap. The model also acknowledges that most individuals undergo behavior relapses after successful change. Knowledge of the model may remove the sense of fatalism many physicians feel when trying to help patients change their behaviors and reinforce the importance of providing the right input at the right stage. For example, repeatedly pressuring a smoker at the precontemplation stage to pick a quit smoking date may create an adversarial relationship, reinforcing the negative behavior and making it less likely the individual will consider cessation. Instead, acknowledging the lack of readiness to quit, spending a few moments to explore the reasons for smoking, and providing education about the harmful health effects of smoking may encourage patient contemplation, setting the stage for eventual cessation.

Issue 5: Failure to Recognize and Acknowledge the Harms of Screening

Clinical prevention saves many lives but also has potential harms, such as complications of diagnostic procedures and patient anxiety. Physicians generally underemphasize the harms of screening in a well-intentioned effort to help as many people as possible. For example, although FOBT has been shown to reduce the relative risk of colorectal cancer death by 33%, the absolute reduction in all-cause mortality associated with testing is only 0.3% and the false-positive rate is high. Many screened individuals must undergo potentially morbid procedures such as colonoscopy to realize the small absolute reduction in mortality. In addition, some physicians continue to offer worthless services due to misguided medicolegal concerns. Patients have also been conditioned by the health care system and the media to believe all preventive care is more beneficial than harmful. These beliefs and practices place patients at an unjustifiably increased risk of harm.

Potential Solutions. When possible, decisions regarding whether to offer a preventive service should be based on absolute reductions in all-cause mortality rather than relative reductions in disease-specific outcomes. Harm counseling must be provided when screening also has the potential for adverse consequences. The informed consent model, developed for clinical research, should be applied. For interventions with proven benefit and minimal or no adverse consequences, such as counseling regarding infant car seat use, informed consent is not necessary.