Health care as if health mattered.
T. Frieden,F. Mostashari
DOI: https://doi.org/10.1001/jama.299.8.950
2008-02-27
Abstract:THERE IS CONSENSUS ABOUT THE NEED FOR FUNDAMENtal change in the US health care system and there has beenattentiontotheimportantproblemsofinadequate access and increasing costs. But the most serious shortcoming—that the nation’s health system is not designed to maximize health—has been overshadowed. Individuals in theUnitedStatesreceiveonlyabouthalftherecommendedmedical services. Only 43% of individuals with diagnosed diabetes, 37% with hypertension, and 25% with hypercholesterolemia have adequate control of their disease; furthermore, less than20%ofsmokerswhotrytoquitreceiveassistancefrom theirphysicians,andonly2%areprescribedpharmacotherapy. Lack of effective primary health care is a public health problem that results in avoidable blindness, amputations, strokes, heart attacks, and premature death. Nearly 9 of 10 Americans with uncontrolled diabetes, hypertension, and hypercholesterolemia already have private or public health insurance. If reforming US health care results only in expanded access to care, costs will increase faster but with limited health benefits. If only cost controls are instituted, even more individuals will be denied access to needed care. Health care must be restructured to make maximizing health the organizing principle. To do this, 3 synergistic changes are needed: (1) payment that offers substantial rewards for disease prevention and effective management of chronic disease; (2) an information system oriented toward prevention; and (3) changes in care management and practice workflows. Efforts to implement each of these changes separately have failed to substantially improve care because they lacked sufficient focus on prevention and because these 3 interventions are needed jointly. Pay-for-performance initiatives have been hamperedby lackof reliable informationonqualityofcareand outcomes. Electronic health records (EHRs) have had limited or no effect because most have not been designed to facilitate, encourage, and trackpreventive services for entirepatientpopulations.Andwithoutfinancial incentivesorthepower of EHRs to track individual patients and patient panels, clinicians are not able to achieve scale or sustain improvements in preventive services or care for those with chronic conditions. Inmostof thecurrentUShealthcare system, treating illness is more profitable than promoting health. Preventive services are usually poorly reimbursed or not reimbursed at all. A patient who has sustained a myocardial infarction may require cardiac surgery that costs $100 000 or more. But a physician who counsels patients to adopt a healthy lifestyle, prescribes the right medications, and follows up regularly to help reduce the patient’s infarction risk can lose money. As a result, health care information systems and practice workflows do not prioritizepreventionor facilitatemanagementofchronicdisease. If a substantialproportionofphysicians’ reimbursementswere based on the proportion of patients who complete all vaccinations and core cancer screenings, have good management of bloodpressureandcholesterol levels, andeitherdonot smoke or have documented assisted attempts at quitting, many more patients would receive these lifesaving preventive services. Todate,pay-for-performancesystemshavegeneratedscant evidence of measurable quality improvements and may offernoadvantageoverfee-for-service inimprovingcareandcontainingcosts. Pay-for-performancesystemsdonotfundamentallyalter thesubstantial financialadvantagesof intensivetreatmentofadvanced illnessoverdiseaseprevention.Currentpayfor-performanceprogramsofferonlysmallperformancebonuses, are often not integrated with quality-improvement initiatives, and often focus on what is easy to measure rather than what willmost improve thehealthofpatients andcommunities. Of thehundredsofambulatorycaremeasuresapprovedbynational quality organizations, none address screening the generalpopulationforhumanimmunodeficiencyvirus,hypercholesterolemia, or depression. Furthermore, these measures often are based on processes performed, such as whether physiciansmeasure lipid levels, rather thanonoutcomes, such as how well lipid levels are controlled. But processes and outcomes may not be correlated. Pay-for-performance might be much more effective at improving health, and would encourage innovation in doing so, if it rewarded outcomes or reliable predictors of outcomes such as control of blood pressure and cholesterol levels and if the rewards were prioritized based on their potential to increase patients’ years of healthy life. Well-designedandeffectively implementedEHRsareanecessary,butnotsufficient,component tochangepaymentstructure and improve health care system performance. EHRs have thepotential toenablevalidclinicalqualitymeasurement,highqualitypreventivecare,andbettermanagementofchronicconditions such as hypertension and hypercholesterolemia. The US Department of Veterans Affairs (VA), which has a financial incentive to improve prevention, implemented an effective EHR system in the mid-1990s and now outperforms the private sector on almost every measure of quality. Patients at VA facilities receive 67% of the recommended level of health care vs 51% received by non-VA patients. A systematic review found that in a small number of institutions, EHRs have