Dolores Flores Ms. Flores has had elevated HbA1c and LDL cholesterol levels for many years, and her blood pressures are generally about 150/95 Never u

Transcription

1 Reducing Cardiovascular Risk: Why is Medical Practice Failing Two-Thirds Thirds of the Time? Thomas Bodenheimer MD Department of Family and Community Medicine University of California at San Francisco This presentation is not an evidence review of cardiovascular risk factor reduction The evidence: JNC 7 [JAMA 2003;289:2560]; update expected 2010 National Cholesterol Education Program, Adult Treatment Panel III (ATP III), updated [Circulation 2004;110:227]; next update expected 2010 Standards of Medical Care in Diabetes-2009 [Diabetes Care 2009;32(suppl 1):S13-S61] S61] We will talk about the limitations of evidence- based medicine and how to address those limitations Objectives By attending this session, participants will Understand the limitations of evidence-based medicine in improving cardiovascular outcomes Learn the components of self-management support for patients with chronic conditions Dolores Flores 52 year old woman with type 2 diabetes, hypertension, elevated cholesterol, and DJD of her knees She is a single mother of 5 children and works in a flower shop She has no health insurance and attends a teaching clinic at SFGH where she has been cared for by a faculty physician, Dr. Lisa Bueno, for the past 6 years 1

2 Dolores Flores Ms. Flores has had elevated HbA1c and LDL cholesterol levels for many years, and her blood pressures are generally about 150/95 Never used tobacco Occasionally checks sugars at home Urine microalbumin elevated Early diabetic retinopathy and LVH on EKG and echocardiogram. Had TIA with history of 20 minutes of right arm weakness and speech difficulty Dolores Flores Dr. Bueno has prescribed Metformin 1000 mg. twice a day Glipizide 10 mg. twice a day Pravastatin 40 mg. once a day Benazepril 40 mg. once a day HCTZ 25 mg. once a day Metoprolol XL 100 mg. once a day Aspirin 81 mg. once a day Tylenol 500 mg. 4 times a day as needed for knee pain Is Ms. Flores being managed with evidence-based medicine? Dr. Bueno does diet/exercise counseling, prescribes proper medications; insulin refused Dr. Bueno orders all routine labs and eye exams listed in practice guidelines; regular foot exams Ms. Flores is being managed with EBM But Dr. Bueno is playing catch-up -- end organ damage has occurred; high risk of disastrous event Dr. Bueno has not assessed Ms. Flores medication adherence; she takes fewer than half of her pills Why didn t evidence-based medicine work for Ms. Flores? Evidence-based medicine is generally conceived as a 2-step process: 1. Research uncovers the evidence 2. Clinicians learn the evidence 2

3 The 2-step process: the results Despite well-designed guidelines for hypertension, hyperlipemia, and diabetes Despite widespread guideline dissemination to physicians for years 65% of people with HBP are poorly controlled 62% with elevated LDL have not reached lipid- lowering goals 63% of people with diabetes have HbA1c > 7 Roumie et al. Ann Intern Med 2006;145:165, Afonso et al. Am J Manag Care 2006;12:589, Saydah et al. JAMA 2004;291:335. The 5-step process 1. Research uncovers the evidence 2. Clinicians learn the evidence 3. Clinicians use the evidence at every visit for every patient 4. Clinicians make sure that patients understand the evidence 5. Clinicians assist and encourage patients to incorporate the evidence into their lives Evidence-based medicine is fine as far as it goes, but it doesn t go far enough How is the US health care system performing on steps 3, 4 & 5? Step 3: Clinicians don t use the evidence at every visit for every patient In a national study of physician performance on 439 process indicators for 30 medical conditions plus preventive care, physicians provided only 55% of recommended care [McGlynn et al. NEJM 2003;348:2635] Step 4: Clinicians often don t make sure that patients understand the evidence A study of 264 visits to primary care physicians using audiotapes Patients making an initial statement of their problem were interrupted by the physician after an average of 23 seconds In 25% of visits the physician never asked the patient for his/her concerns at all Marvel et al. JAMA 1999;281:283 3

4 Step 4: Clinicians often don t make sure that patients understand the evidence Most physicians blame their patients for medication nonadherence [O Brien et al. Medical Care Review 1992;49:435] However, only 37% of patients in one study were adequately informed about medications they were taking [Roter and Hall. Ann Rev Public Health 1989;10:163. In one study 76% of patients with type 2 diabetes received limited or no diabetes education [Clement, Diab Care 1995;18:1204] Audiotaped study: in most visits, primary care physicians provide inadequate explanations of meds [Tarn et al, Arch Int Med 2006;166:1855] Step 4: Clinicians often don t make sure that patients understand the evidence Old audiotaped study of 336 medical with 34 physicians, doctors devoted an average of 1.3 minutes to giving information The docs thought they gave 8.9 minutes 88% of the information was in technical language Waitzkin, JAMA 1984;252:2441. Step 4: Clinicians often don t make sure that patients understand the evidence The 50% rule Asking patients to repeat back what the physician told them, half get it wrong. [Schillinger et al. Arch Intern Med 2003;163:83] Asking patients: Describe how you take this medication -- 50% don t understand and take it differently than prescribed [Schillinger et al. Medication miscommunication,, in Advances in Patient Safety (AHRQ, 2005)] 50% of patients leave the physician office visit without understanding what the physician said [Roter and Hall. Ann Rev Public Health 1989;10:163] Failure to provide information to patients about their chronic condition is associated with unhealthy behaviors. If people don t know what to do, they don t do it. [Kravitz et al. Arch Intern Med 1993;153:1869. O Brien et al. Medical Care Review 1992;49:435] Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives Giving information to patients by itself does not improve clinical outcomes; an additional factor is needed That additional factor appears to be collaborative decision making, which makes the patient an active participant in his/her management While half of patients surveyed want to leave final decisions to their physician, 96% want to be offered choices and to be asked their opinions [Levinson et al. JGIM 2005;20:531]. 4

5 Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives In a study of 1000 physician visits, the patient did not participate in decisions 91% of the time [Braddock et al. JAMA 1999;282;2313] African-American American and other ethnic minority patients visiting a white physician report less involvement in medical decisions, less partnership with the physician, and lower levels of trust in physicians. [Cooper et al. Ann Intern Med 2003;139:907] Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives Evidence on collaborative decision-making is fairly persuasive A participatory relationship between physician and patient is associated with patient adoption of healthy behaviors [O Brien et al. Medical Care Review 1992;49:435]. Collaborative care, which encourages patients to be active participants, improves the chances that the patient is in agreement with the decisions made and thereby improves health-related behaviors [Mead and Bower, Patient Educ Couns 2002;48:51; Norris et al. Diabetes Care 2001;24:561] Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives One needs both information-giving and participatory decision making to improve healthy behaviors In a study of 752 ethnically diverse patients, more effective patient-provider provider communication (both information giving and collaborative decision making) was associated with healthier self-reported behaviors, medication use, dietary adherence, and exercise [Piette JD et al. J Gen Intern Med 2003;18: ]. 633]. Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives 20-minute pre-visit meeting to prepare patients with diabetes to participate in decision-making Pre-activated patients participated more (shown by audiotapes) than control patients Pre-activated patients had better HbA1c levels than control patients Greenfield, Kaplan et al. J Gen Intern Med 1988;3:448 5

6 Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives Patients with diabetes supported by their providers to have some control over health care decisions had lower HbA1c levels than those without such control [Williams et al. Diabetes Care 1998;21:1644] Patients more actively involved in their care (medication use, glucose testing, diet, exercise and foot care) had better HbA1c levels than those less involved [Heisler et al. Diabetes Care 2003;26:738] The greater the amount of patient participation in the medical visit, the higher the level of medication adherence [O Brien et al. Medical Care Review 1992;49:435] Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives Heisler et al. found significant associations between improved information giving by the physician, more participatory decision making, enhanced self-efficacy, efficacy, healthier behaviors, and better outcomes in patients with diabetes. They conclude that Enhancing patient-provider provider communication and shared decision making have been shown to result in greater patient satisfaction, adherence to treatment plans, and improved health outcomes The consistency of these studies findings of improved physiologic outcomes and reported health status is impressive. [Heisler et al. J Gen Intern Med 2002;17:243] Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives For optimal management of their chronic conditions, patients need Information, and To be active participants in their care The Chronic Care Model Widely accepted as the guide to improving chronic care Its message: the goal is the informed activated patient and family Ms. Flores -- uninformed passive patient or informed activated patient? Ms. Flores did not have time in her life to learn about her CVD risk factors Dr. Bueno did not have time in the 15 minute visit to teach Ms. Flores Ms. Flores seldom asked questions in visits to Dr. Bueno because she know Dr. Bueno was very busy 6

7 Ms. Flores -- uninformed passive patient or informed activated patient? Health was a low priority in Ms. Flores life. She had a job and 5 children to manage, and little money and a minimal support system She was very active in her life, but passive as a patient She was unable to do much physical activity due to DJD Patients with diabetes in chronic pain have more trouble following a diet plan, exercise plan, and poorer adherence to medications [Krein et al. Diabetes Care 2005;28:1534] Evidence-based medicine steps 3, 4, & 5 are not performed in most visits Step 3: Clinicians don t use the evidence at every visit for every patient. Process measures performed only 55% of the time. Step 4: Clinicians often don t make sure that patients understand the evidence. The 50% rule Step 5: Clinicians assist and encourage patients to incorporate the evidence into their lives. Collaborative decision-making takes place in only 9% of decisions Evidence-based medicine steps 3, 4, & 5 are not performed in most visits Who is falling down on the job? Is it a patient problem? A doctor problem? A system problem? You can t fix it unless you know 7

8 Is it a patient problem? We cannot blame patients for failing to adhere to medical advice if They do not understand what the physician expects them to do They were never asked if they agree with what the physician expects them to do If we do our job well -- information giving and collaborative decision-making -- and patients do not choose to make the life changes needed to bring their chronic condition under control, then the responsibility lies with the patient Is it a system problem? The 15-minute visit Primary care visit lengths have increased by 2 minutes in the past 10 years But required work has doubled or tripled The main microsystem problem: the 15-minute visit Race-discordant visits (patient and physician from different racial/ethnic group, usually white physician and minority patient) are on average 2.2 minutes shorter than race-concordant concordant visits Even though minority patients on average have more health problems Tyranny of the urgent Acute problems crowd out time for routine management of chronic illness and preventive care Wagner et al. Milbank Quarterly 1996;74:511. 8

9 The main system problem: the 15-minute visit Hamster Care Across the globe doctors are miserable because they feel like hamsters on a treadmill. They must run faster just to stand still The result is a reduction in the quality of care and an increase in burnout among doctors. [Morrison and Smith, BMJ 2000;321:1541] Funny things doctors write in their charts when they are working like hamsters: She is numb from her toes down. Patient has two teenage children, but no other abnormalities. 9

10 While in ER, she was examined, X-rated and sent home. The main problem: the 15-minute visit Primary care practices with longer visit times scored better on quality indicators for diabetes, asthma and coronary heart disease High-volume primary care physicians had visits 30% shorter than low-volume physicians. High-volume: fewer preventive services; poorer measures of patient satisfaction and physician-patient patient relationship. Active patient participation in the visit tends to increase visit length -- in one study by 34% Campbell et al. BMJ 2001;323:784. Zyzanski et al. J Fam Pract 1998;46:397. Dugsdale et al. J Gen Intern Med 1999:14(supple 1):S34. Hornberger et al. J Gen Intern Med 1997;12:597 The main system problem: the 15-minute visit Physicians cannot be expected to provide all acute, chronic and preventive care in the 15-minute visit The rushed 15-minute visit may be largely responsible for Doctors interrupting patients Poor information giving Rare collaborative decision making Primary care: Impossible to do it right A primary care physician with an panel of 2500 average patients will spend 7.4 hours per day doing recommended preventive care [Yarnall et al. Am J Public Health 2003;93:635] A primary care physician with an panel of 2500 average patients will spend 10.6 hours per day doing recommended chronic care [Ostbye et al. Annals of Fam Med 2005;3:209] 10

11 Back to the original questions: Why didn t evidence-based medicine work for Ms Flores? Why doesn t it work for many people? 65% of people with HBP are poorly controlled 62% with elevated LDL have not reached lipid-lowering lowering goals 63% of people with diabetes have HbA1c > 7 Who is falling down on the job? Sometimes patients Sometimes doctors Usually the system What can primary care do to encourage the incorporation of evidence-based medicine into the lives of patients with chronic conditions or risk factors? What is self-management? Self-management is what people do every day: decide what to eat, whether to exercise, if and when they will monitor their health or take medications. Everyone self-manages; the question is whether or not people make decisions that improve their health-related behaviors and clinical outcomes. People who are motivated to make daily decisions and choose actions favoring healthy behaviors are sometimes called good self-managers. Bodenheimer et al. Helping Patients Manage their Chronic Conditions. California Healthcare Foundation,

12 What is self-management support? Self-management support is what health caregivers do to assist and encourage patients to become good self- managers. Institute of Medicine definition: the systematic provision of education and supportive interventions to increase patients skills and confidence in managing their health problems, including regular assessment of progress and problems, goal setting, and problem-solving support. IOM. Priority Areas for National Action: Transforming Health Care Quality. Washington DC: National Academies Press, 2003, p 52. The components of self-management support Provide information Intensive skills training (disease specific) Encouraging healthy behavior change Teach patients problem-solving skills Assisting patients with psychosocial issues and the emotional impact of having a chronic condition Provide ongoing and regular follow-up Encourage and train patients to become active participants in their care Providing information Patient education on diabetes improves patient knowledge, but does not improve glycemic control 59 trials of hypertension management: patient education alone does not work Norris et al. Diabetes Care 2001;24:561 Fahey et al. Cochrane Review 2005; Jan 25;(1):CD Providing information Cochrane review of 12 trials on asthma:patient education alone does not improve outcomes nor frequency of asthma-related ED visits [Gibson et al. Cochrane Review 2002;(2):CD001005] Cochrane review of arthritis patient education alone: no long term benefits for adults with rheumatoid arthritis [Riemsma et al. Cochrane Review 2003;(2):CD003688] Interventions to improve medication adherence, education alone had no effect [Haynes et al. Cochrane Review 2002;(2):CD000011] 12

13 Providing information Information-only only patient education is necessary but not sufficient to achieve improved outcomes The components of self-management support Provide information Intensive skills training (disease specific) Encouraging healthy behavior change Teach patients problem-solving skills Assisting patients with psychosocial issues and the emotional impact of having a chronic condition Provide ongoing and regular follow-up Encourage and train patients to become active participants in their care Intensive skills training Hypertension Intensive skills training Atrial Fibrillation Meta-analysis analysis of 18 RCTs Patients who self-monitor their blood pressure at home have better blood pressure control than people with usual care (blood pressures taken only during a medical visit). Cappuccio et al. BMJ 2004;329:145 Compared with usual care, patients who self-monitor and self-adjust their warfarin doses at home have INR values more frequently in the target range. [Sawicki. JAMA. 1999;281:145] Literature review of warfarin self-monitoring: Home self-monitoring is more effective than usual care and as effective as care through specialized anticoagulation clinics for keeping INRs within a therapeutic range. Appropriate patient selection and thorough patient training are essential. [Yang et al. Am J Hematol. 2004;77:177] 13

14 Intensive skills training Diabetes Home glucose testing is questionably associated with improved glycemic control in patients with type 2 diabetes [Davis, Diabetes Care 2006;29:1764; Cochrane Review 2005;CD005060] For insulin-dependent type 2 diabetes, patients self-administering their insulin based on algorithm had better glycemic control than physician-managed insulin, with no difference in hypoglycemic episodes. [Davies et al. Diabetes Care 2005;28:1282; Davidson et al. Am J Med 2005;118(suppl 9A):27S] The components of self-management support Provide information Intensive skills training (disease specific) Encouraging healthy behavior change Teach patients problem-solving skills Assisting patients with psychosocial issues and the emotional impact of having a chronic condition Provide ongoing and regular follow-up Encourage and train patients to become active participants in their care Goal-setting and action plans Patient chooses goal: to lose weight Unrealistic action plan: I will lose 20 pounds in the next month. I will walk 5 miles a day. Realistic and specific action plan: I will eat one candy bar each day rather than the 5 per day I eat now. I will walk for 15 minutes each day after lunch. Self management support If people don t want to do something, they won t do it Kate Lorig RN, Dr. PH Stanford Medical School 14

15 Goal-setting and action plans Goal-setting and action plans Ammerman et al. reviewed 92 studies involving behavioral interventions to improve diet. Goal setting was associated with a greater likelihood of obtaining a significant intervention effect for 3 outcomes (less total fat, less saturated fat, and more fruits/vegetables). Ammerman et al. Preventive Medicine2002;35:25. Cullen reviewed 13 studies utilizing goal-setting in adult nutrition education. Persons engaged in goal setting to improve diet did better in terms of self-reported dietary change, weight loss and improved serum cholesterol than control groups. Cullen et al. J Am Diet Assoc 2001;101:562. Goal-setting and action plans In 2004, Shilts reviewed 28 studies of goal-setting for dietary and physical activity behavior change. 32% of the studies were evaluated as fully supporting the use of goal setting. The review concluded that goal setting has shown some promise in promoting dietary and physical activity behavior change among adults Shilts et al. Am J Health Promotion 2004;19:81. There is no improvement, Henry. Are you sure you ve given up everything you enjoy? 15

16 Goal-setting and action plans A review of the evidence on improving diet, published by the Agency for Healthcare Research and Quality, included goal-setting in a list of a few intervention components shown to be associated with improved behavioral outcomes. [Systematic Evidence Review Number 18. Counseling to Promote a Healthy Diet. AHRQ April 2002] Pignone et al. reviewed dietary counseling for the USPSTF. Goal-setting was one element associated with improved outcomes [Am J Prev Med 2003;24:75] Goal-setting and action plans The American Diabetes Association website s guide to changing habits is entitled Setting Goals Helps You Take Charge of Diabetes. The guide suggests making a specific and realistic action plan, for example, walk for half an hour 3 times a week [ In three separate statements of standards for diabetes education, the American Association of Diabetes Educators recommends that diabetes education should include goal-setting [ The American Heart Association scientific statement on treating obesity- related heart disease risk factors recommends self-monitoring, goal- setting, stress management and social support as behavioral strategies for improving diet and physical activity. [ The components of self-management support Provide information Intensive skills training (disease specific) Encouraging healthy behavior change Teach patients problem-solving skills Assisting patients with psychosocial issues and the emotional impact of having a chronic condition Provide ongoing and regular follow-up Encourage and train patients to become active participants in their care Follow-up: diabetes Cochrane Review (Griffin and Kinmouth): patients with diabetes who had regular follow-up had better HbA1c levels than without such follow-up [Griffin and Kinmonth. Cochrane Review 2000;(2):CD000541] Norris et al. meta-analysis: analysis: the benefits of self- management for patients with diabetes diminishes over time; sustained regular follow-up is needed. Total time spent with a patient is closely correlated with improved glycemic control [Diabetes Care 2002;25:1159] 16

17 Follow-up: hypertension, CHF A review of 59 trials of hypertension management. Regular follow-up was essential to improving blood pressures Fahey et al. Cochrane Review 2005; Jan 25;(1):CD CHF: meta-analysis analysis of 30 trials. Patients with regular follow-up had 30% reduction in CHF admissions and 20% reduction in all-cause mortality. Holland et al. Heart 2005;91:899. Tentative conclusions about self- management support Information is necessary but not sufficient to improve chronic disease outcomes; in addition, patients need to be active participants in the management of their conditions Patients need to learn self-management skills. Self- monitoring (blood sugars, blood pressures, asthma symptoms, warfarin doses) is most successful if patients learn how to react to a measurement. Tentative conclusions from this evidence The triad of goal-setting, action-planning and problem-solving, while not rigorously evidence-based, appears to be central to improving chronic disease behaviors and outcomes Regular and sustained follow-up is critical to any chronic disease management or prevention program How might Dr. Bueno s primary care practice transform itself to improve the care of patients with cardiovascular risk factors? 17

18 Teams or teamlets Physicians cannot do much self-management support. We don t have time. We need to understand it and encourage it but someone else needs to do it At SFGH we have health coaches (medical assistants or health workers) provide self- management support. Other primary care practices have RNs, health educators. Other patients (peer coaches) can do it: promotora programs Teams or teamlets The health coach plus physician are called the teamlet Called teamlet because It is part but not all of the team It is small The coach helps patients and also helps physicians by offloading time-consuming discussions with patients Coaches are ethnic-language concordant with their patients How does the teamlet model allow the 5-step evidence-based medicine process to work? 3. The care process is transformed so evidence is used at every visit for every patient Routine evidence-based processes (labs, foot exams, eye exams, patient education) 4. The health coach makes sure that patients understand the evidence (the 50% rule) Coaches close the loop 5. The teamlet assists and encourages patients to incorporate the evidence into their lives Coaches set goals collaboratively with patients. They do medication reconciliation and work on barriers to adherence. Conclusion Who is responsible for the poor control of Dolores Flores cardiovascular risk factors? If the physician did not make sure Ms. Flores had the information she needed If the physician did not collaborate with Ms. Flores in decisions about her care If these things didn t happen because it was impossible in the 15-minute visit Then it s a system problem Teams or teamlets are one way to address this system problem and perform the entire 5-step evidence-based medicine process 18