EDITORIAL

American Health Care, and Notably Cardiovascular Care, Cannot be Defined by WHO Mortality Rankings

Gerald Dorros, MD, ScD and Adam J. Pressman, MBA

Gerald Dorros, MD, ScD and Adam J. Pressman, MBA

The costly American health care system, allegedly, has overpromised and underperformed, and, as such, a reduction in public and private funding for research and physician compensation seems, to some, justified. This perspective, the focus of intense political debate, has been erroneously advanced citing the US’s poor WHO (World Health Organization) mortality rankings and that higher ranking countries spend considerably less per capita for health care. This mistaken allegation, by aligning flawed WHO mortality statistics and country rankings, and per capita country health care expenditures, should be seen as prima facie evidence of American health care inferiority and incompetence, and, by inference, poor physician performance. These unfortunate statements, pegged to blemished data and invalid comparisons, have only fomented confusion, tension, turmoil, and angst among Americans toward health care providers, and have diverted focus and energy away from the crucial issues necessary to be addressed in the attainment of improved health care delivery. 

The most significant challenges facing any individual or body in assessing a problem remain data comparability, which demands comprehension of definitions, data collection methodology, accuracy, and verification, whether employing external or internal benchmarks; unfortunately, the WHO states that its data collection, methodologies, verification, and accuracy are seriously defective, which sharply contrasts to those regarding US national vital statistics. As such, the methodology of using internal benchmarking, when considering US national vital statics, e.g., cardiovascular statistics, consistently collected over time, could provide a better assessment of US health care. In 2006, cardiovascular disease accounted for 34% of all deaths (831,272/2,426,264), and was mentioned in 56% (1,347,000) of all US deaths.1 If these values were compared to those from a prior time period, perhaps, a different perspective of the effect of the US’s cardiovascular health care upon longevity might be reached.  A positive conclusion would reflect that Cardiology’s transparent and relentless pursuit of better patient care through research (meticulously detailing, defining, and verifying study metrics and data), education, technologically advanced diagnostic methods, novel therapies, and open collegial dialogue, while meticulously detailing, defining, and verifying study metrics and data, resulted in improved patient outcomes. Such a finding would help validate the medical community’s pursuit of as well as attainment of better health care delivery. 

In the US, all vital events, i.e., births, deaths, terminations of pregnancy, marriages, and divorces are registered locally within the 57 vital registration jurisdictions, which do direct statistical reporting to the National Center for Health Statistics (NCHS).  Each of the jurisdictions is also a member of NAPHSIS, the National Association for Public Health Statistics and Information Systems, which was formed in 1933. NAPHSIS sets the national standards for data collection, exchange, issuance, and verification and electronic systems to collect, maintain, and disseminate records and data in a way that is timely, accurate, and protects confidentiality and security.  The task of civil registration and recording of the over 11 million annual vital events is enormous, among them being the 4.3 million births, and 2.4 million deaths, which are reported directly to state and local vital statistics offices in the form of certificates of birth, death, marriage, divorce, etc. Each record is unique, and must be logged-in and evaluated for data quality and assigned file numbers; these data records serve two purposes: 1) the legal registration of the record; and 2) to provide certified copies of the vital data, such as birth and death certificates. Finally, these vital event data, derived from the certificates filed with each jurisdiction, become the vital statistics of that jurisdiction, and become part of the nation’s vital statistics through a cooperative agreement with NCHS, the Vital Statistics Cooperative Program (VSCP) and with NCHS to provide data that meet the NCHS national standards for consistency, quality, and timeliness.2 Standard forms for the collection of these data, and model procedures for the uniform registration of the events, including collection, coding, editing, transmitting original data, as well as an automated system for coding and classifying cause-of-death information from death certificates, enables the National Vital Statistics System, part of the CDC’s National Center for Health Statistics, to publish the national vital statistics.3

The preceding sharply contrasts with that of the United Nation’s WHO, which desires accurate information,4 but acknowledges the fact they have not been able to achieve accurate, verifiable, and consistent data. For example, the WHO states, regarding their under 5-years mortality statistics, “Even though many countries have collected information on child mortality in recent years, the high demand for very recent child mortality trend information is difficult to meet through household surveys. High quality of vital registration systems (completeness of registration) and high quality of survey or census data collection are crucial — WHO does estimate the level of underestimation of vital registration systems and there clearly is substantial variation in data quality and consistency across countries.”5

WHO openly states that the assumption of accuracy of countries’ supplied mortality statistics is inappropriate and is seriously flawed. By rank, in 2004, the 5 leading causes of all adult deaths in the US were: 1) heart disease (26%); 2) malignancy (23%); 3) stroke (6%); 4) chronic lower respiratory diseases (5%); and 5) accidents (unintentional injuries) (5%), which accounted for 65% of American deaths, and are: 1) similar to those of Canada6 and Great Britain;7 but 2) differ sharply from  the rankings of WHO and the US Census Bureau’s International Database for Leading Causes of World Mortality in 2002, which found: 1) communicable diseases, maternal and perinatal conditions, and nutritional deficiencies (32% of deaths); 2) infectious and parasitic diseases (20%); 3) tuberculosis (3%); 4) sexually transmitted diseases/HIV/AIDS (5%); and 5) diarrheal diseases (3%). These dissimilar ranking categories, regardless of data accuracy, by themselves preclude any meaningful comparative analysis. The misinterpretations of these disparate mortality statistics resulted in the non-sequitur that the “excessive” amounts of money spent on health care had produced “inferior results” as evidenced by world mortality rankings.

WHO acknowledges that, “according to data provided by 112 Member States… (only) one-third of the estimated (annual) 56 million deaths are recorded in death registration systems.” In the 2010 WHO indicator code book, “there is a dearth of data on adult mortality, notably in low-income countries. Methods to estimate adult mortality from censuses and surveys are retrospective and possibly subject to considerable measurement error.” In addition, regarding age-standardized mortality rate, “cause-of-death distributions are estimated from death registration data, and data from population-based epidemiological studies, disease registers, and notification systems for selected specific causes of death. Causes of death for populations without usable death-registration data were estimated using cause-of-death models together with data from population-based epidemiological studies, disease registers, and notification systems for 21 specific causes of death… uncertainty in estimated all-cause mortality rates ranges from around ±1% for high-income countries to ±15-20% for sub-Saharan Africa, reflecting large differences in the availability and quality of data on mortality, particularly for adult mortality. Uncertainty ranges are generally larger for estimates of death rates from specific diseases. For example, the relative uncertainty for death rates from ischemic heart disease ranges from around ±12% for high-income countries to ±25-35% for sub-Saharan Africa. The relatively large uncertainty for high-income countries reflects a combination of uncertainty in overall mortality levels, in cause-of-death assignment, and in the attribution of deaths coded to ill-defined causes.”8 Thus, any country-to-country comparative data analyses are inaccurate, at best.

An estimated 81,100,000 American adults (more than 1 in 3 adults) have one or more cardiovascular disease (CVD) type, of which nearly half are >60 years, with high-blood pressure (HBP) in 74,500,000, coronary heart disease (CHD) in 17,600,000, myocardial infarction (heart attack) in 8,500,000 (with angina pectoris [chest pain] in 10,200,000; heart failure in 5,800,000), and stroke in 6,400,000. Between 1994 and 2007, the Healthcare Cost and Utilization Project, from the Agency for Healthcare Research and Quality, detailed the national risk-adjusted mortality rate decreases for acute myocardial infarction (AMI), congestive heart failure (CHF), and stroke, ranging from 60% among CHF inpatients to 35% among stroke inpatients.9 Two sets of mortality statistics (overlapping time periods, 1996-2006 and 2000-2007) showed continuing improvements:

  1. In 2006, CVD claimed 831,272 lives (34% of all deaths), of which 151,000 were <65 years old. From 1996 to 2006, death rates from CVD declined 29%, while actual CVD deaths declined 13%.
  2. In 2006, CHD caused 425,425 deaths, yet the CHD death rate declined 36.4%, and the actual number of deaths declined 22%. From 2000-2007, the death rate for patients treated in-hospital for AMI went down 36% (106 to 67 deaths per 1000 admissions). Mortality among Medicare hospitalizations for AMI decreased by 38% (106 to 66 deaths per 1000 admissions), Medicaid hospitalizations decreased by 27% (103 to 75 deaths per 1000 admissions), with similar decreases recorded in the uninsured and privately insured.10
  3. In 2006, CHF caused death in 60,337 of 282,754 CHF patients, accounting for 7% of CVD deaths. From 2000-2007, the death rate fell 49% (55 to 28 deaths per 1000 admissions), for men 52% and women 46%, and for Medicare 51% (55 to 27 deaths per 1000 admissions). 
  4. In 2006, stroke killed 137,119 people. From 1995-2005, the stroke death rate fell 34% with actual stroke deaths declining 18% (125 to 92 deaths per 1000 stroke admissions) and from 2000-2007 decreasing 35% (143 to 92 deaths per 1000 stroke admissions).   

The improved CVD mortality statistics correlate with increasing usage of evidenced-based medicine data, and indicated interventional (e.g., angioplasty) or surgical (coronary artery bypass surgery) procedures, and a resulting increased life expectancy.

Nevertheless, the crystal clear improvements seen in cardiovascular disease data do not seem to correlate with overall US mortality statistics and mean life expectancy values, because these data points fail to acknowledge the significant non-medical social issues, which increase mortality statistics and decrease life expectancy values. In the US, accidents (unintentional injuries) are the leading cause of death for all races and sexes, ages 15-44 years. The incidence varies with age (15-19 years, 49% of deaths; 20-24 years, 45%; 25-34 years, 35%; 35-44 years, 21.1%). In addition, for ages 15-34 years, the second and/or third causes of death were homicide (18%, and 11%) and suicide (11%, and 12%).11 These social issue causes, which accounted for >70% of deaths for ages 15-24 years, and for 55% of deaths for 25-34 years, more than likely occurred before any health care professional interacted with these individuals. These issues, as well as the effect of smoking and obesity, which are not addressed herein, adversely impact mortality and longevity statistics, and led health care economists Robert Ohsfeldt and John Schneider in The Business of Health12 to conclude that when taking these social issues into account, the US, when viewing overall mortality rates and rankings, rose to the leading world position with the best life expectancy value.13 

During the last four decades, well-educated and intellectually curious physicians, research scientists, and design and biomedical engineers, in conjunction with medical device, imaging, and pharmaceutical companies, have produced major medical diagnostic and therapeutic advances. These novel devices, drugs, and therapies enabled physicians to provide not only the best treatments but demonstrated continued improved patient outcomes and reduced mortality rates. American life expectancy is significantly affected by cardiovascular diseases, but the significant social issues that directly influence US mortality statistics and life expectancy, killing too many Americans before their time, not only require attention, but also assessment and a corrective-action plan. Thus, a superficial glance at a few numbers out of context should not be used to glibly impugn physicians, health care professionals, or our health care system.

While mortality is a definable endpoint, mortality may not or should not necessarily define a therapy, or a health care system, and certainly, disparate and noncomparable mortality statistics have no place in such assessments. Critiques, criticism, and open, passionate debate of everyone and everything are necessary for medicine’s evolution. US health care expenditures have produced very good results, yet Americans still can benefit from focused fiscal management, which can use “saved” dollars on primary and tertiary health care services, and research rather than edifices and administrators, and a comprehensive health insurance, which should facilitate patient access, and, as a result, better outcomes. Governmental regulatory requirements and oversight are critical elements in the path to attain better health care; however, this should not be construed to imply that, within the present regulated environment, health care professionals have not produced or delivered. The internal benchmarking of cardiovascular statistics has shown that they have delivered. Unfortunately, the hubris of politicians, a territorial hazard and occupational trait, will cause them to 1) squander the government’s limited resource of competence; and 2) micromanage health care, which will result in grandiose pledges and expected underperformance. As a result, more direct and bureaucratic governmental controls, oversight, mandates, and impositions upon health care providers, hospitals, and the medical industry will slow rather than enable the necessary continued advances in American health care.

References

  1. www.americanheart.org/downloadable/heart/1265665152970DS-3241%20HeartStrokeUpdate_2010.pdf
  2. This paragraph paraphrases Steven Schwartz, PhD, Registrar and Assistant Commissioner, Bureau of Vital Statistics New York City Department of Health and Mental Hygiene: www.naphsis.org/NAPHSIS/files/ccLibraryFiles/Filename/000000000676/NAS%20VS%20Workshop-%20State%20Perspective-%204-23-08.pdf
  3. www.cdc.gov/nchs/nvss/about_nvss.htm
  4. www.who.int/healthmetrics/tools/Version_4.00_Assessment_Tool3.pdf
  5. www.who.int/healthinfo/statistics/indunder5mortality/en/
  6. www.statcan.gc.ca/daily-quotidien/101130/dq101130b-eng.htm
  7. www.statistics.gov.uk/pdfdir/dth1010.pdf
  8. www.who.int/whosis/indicators/WHS10_IndicatorCompendium_20100513.pdf
  9. www.hcup-us.ahrq.gov/reports/statbriefs/sb98.pdf
  10. www.hcup-us.ahrq.gov/reports/statbriefs/sb98.pdf
  11. www.cdc.gov/nchs/data/nvsr/nvsr56/nvsr56_05.pdf
  12. www.aei.org/book/859
  13. blogs.wsj.com/numbersguy/does-the-us-lead-in-life-expectancy-223/

____________________________________

Dr. Dorros is the Medical Director of the William Dorros-Isadore Feuer Interventional Cardiovascular Disease Foundation, Ltd., in Wilson, Wyoming and a trustee of the Hudson Institute, Washington, DC. Mr. Pressman is a principal of AT Kearny Consulting, Chicago, Illinois.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors reported no conflicts of interest regarding the content herein.
Manuscript submitted April 11, 2011, provisional acceptance given May 25, 2011, final version accepted August 3, 2011.
Address for correspondence: Gerald Dorros, MD, Medical Director, The William Dorros-Isadore Feuer Interventional Cardiovascular Disease Foundation, Ltd., P.O. Box 1654, 1120 South Thunder Road, Wilson, WY 83014. Email: gdorros@dorrosfoundation.org