Measuring health outcomes
Health outcome measurement quantifies changes in health status associated with providing or withholding healthcare. It requires a clear account of the expected benefit, a suitable metric and observations that establish the direction and size of change.
Health and its boundaries
Health lacks a universally agreed meaning despite the standardised units used for physical parameters such as weight, blood pressure, lung function, nerve signal transmission, cardiac output and blood chemistry. The World Health Organization's 1946 Constitution defines health as complete physical, mental and social well-being, rather than merely the absence of disease or infirmity. This leaves questions about what well-being means and when it is complete.
A health economist's 2018 critique was that the definition's three facets did not accommodate cultural and ethnic factors: traditional Chinese medicine, for example, recognises signs and symptoms that physicians trained elsewhere may discount. The critique also identified changing social organisation, practices and values as sources of new health-related concerns, including contraceptive practice, gender identity, cosmetic appearance and domestic violence. Broadening assessment to quality of life, well-being, life satisfaction or happiness did not, in that account, resolve the difficulty of defining the concept or establishing a standard method of observing it. For measurement, the practical proposal was to treat health as quantifiable while recognising that no single approach had become the standard.
What counts as a health benefit depends on these boundaries. Relief from arthritic pain is readily classified as a health benefit, whereas benefits from cosmetic tattoo removal may be regarded as largely unrelated to health. Older people whose health and social care needs overlap also illustrate this boundary.
Expected benefits and observed outcomes
Healthcare interventions generally aim to improve the expected trajectory of health compared with what would occur without them. Expected benefits include saving lives, relieving symptoms, maintaining or improving function and potentially extending life through early detection. The question of whether interventions work applies both to individual patients and to populations.
Health outcome measurement can be defined as quantifiable change in health status resulting from providing or withholding healthcare. It supplies three pieces of information: whether anything changed, the direction of change and its magnitude. Quantifiable evidence of expected treatment benefits should be available to healthcare professionals and patients.
Measurement begins with description. Disease staging systems such as TNM and Dukes classification in colon cancer describe disease rather than the whole patient experience and are therefore only partial indicators of health. Clinicians may assess health through familiar clinical parameters while patients judge their condition differently. In the health economist's account, neither perspective takes precedence over the other.
A target metric and repeated observations
The rationale for intervening should guide selection of a target metric that the intervention is expected to influence. Tracking that metric over time produces evidence of the direction and magnitude of change. For example, a planned weight loss programme may produce measurable changes in standardised units. The patient is weighed before and after the intervention using the same weighing machine, and the difference is calculated. The sign gives the direction of change and the arithmetic difference gives its magnitude.
A health status measurement at one time point does not itself establish an outcome. Repeated before-and-after observations allow a change in health status to be inferred.
Recording outcomes and using the results
Systematic recording of patients' status after treatment has a long history. Florence Nightingale's nineteenth-century classification for patients leaving her care distinguished those who were relieved (better), unrelieved (unchanged) or dead. In the early twentieth century, Ernest Codman promoted outcome measurement through follow-up long enough to determine whether treatment had permanently relieved the presenting condition or symptoms.
Health outcome measurement is an integral part of clinical studies and informs the planning, financing, management and delivery of healthcare. It can guide pharmaceutical investment decisions and should provide information for patients and other healthcare consumers. Within routine care, outcome information can help refine decisions about preferred treatments; its absence permits practices of unproven benefit to continue.