Health Statistics and Epidemiology

“Health statistics represent people with the tears wiped off.”
–Sir Austin Bradford Hill (1897-1991), Pioneer of the randomized clinical trial

Statistical concepts and methods are essential for both biomedical and health services research and both health statistics and epidemiology make abundant use of their analytic strategies. For population-based surveys, such as those conducted by the National Center for Health Statistics, the methods used were pioneered by social scientists--and social epidemiology is of increasing importance.

Counting is the “stock in trade” of these endeavors but investigators, clinicians, and managers of all disciplines need to remember that not all that “counts” can be counted.  Caring, compassion, and hope--the foundation of therapy and healing--cannot be easily counted or measured. Nevertheless, familiarity with health statistical, epidemiological, and survey applications and methods is now a requirement for all medical graduates.
Of far greater importance is the internalization of the underlying concepts of statistical reasoning such as: bias (in its many forms), randomization, probability, sensitivity and specificity--the distinction between efficacy and effectiveness--and above all, the difference between statistical and clinical significance.

Counting the dead initially gave way in the 17th century to recording the “causes” of death, and over the decades, to expanding and diversifying their “labels” based on notions of causality. These too have changed frequently over time, thus giving us the 10th edition of the International Classification of Diseases (ICD-10).  To address the need for studying the natural history of initial symptoms, their mutation into disease, and its early prevention, the International Classification of Primary Care was created. Accurate and honest labeling is essential for the production of useful epidemiological and statistical information.  However there still remains doubt about the reliability and validity of some morbidity and mortality survey data.

Epidemiologists initially investigated the origins of infectious disease outbreaks.  Now they also focus their efforts on chronic illness, behavioral and societal disorders, and evaluating clinical and population-based interventions and services.   Epidemiology developed in medical schools in the 19th century and all epidemiologists were physicians.  Then in the early 20th century, schools of public health were created and assumed the primary responsibility for developing modern epidemiology. The pendulum has swung back and most medical schools now have clinical epidemiology units or departments that develop and use epidemiological concepts and methods to study disease, assess treatments and monitor care in hospital, ambulatory and community settings. These venues provide alternatives for curious young physicians to investigate health problems and services outside of the biomedical laboratories; the demand for such skills continues to grow.