Measures of Mortality

PRINCIPLE OF EPIDEMIOLOGY MEASURES OF MORTALITY Trend in observed numbers of cancer deaths for men and women in the United States 1969–2014 2 3 Rates are per 100,000 and age-adjusted to the 2000 US standard population 4 5 Major causes of death in children under 5 years in 2015 MORTALITY RATES 6 Remember … 7  the population changes over time,    the number of persons in the population at midyear is generally used as an approximation of average population. anyone in the group represented by the denominator must have the potential to enter the group represented by the numerator. We may not always be interested in a rate for the entire population; perhaps we are interested only in a certain age group, in men or in women, or in one ethnic group. age-specific mortality rate 8 specific rate 9    In putting a restriction on age the same restriction must apply to both the numerator and the denominator, so that every person in the denominator group will be at risk for entering the numerator group. When such a restriction is placed on a rate, it is called a specific rate. We could also place a restriction on a rate by specifying a diagnosis, and thus limit the rate to deaths from a certain disease, that is, a disease-specific or a cause-specific rate. disease-specific rate 10 restrictions on more than one characteristic 11 Time 12    Time must also be specified in any mortality rate. Mortality can be calculated over 1 year, 5 years, or other. The period selected is arbitrary, but it must be specified precisely. CASE-FATALITY 13 CASE-FATALITY 14      what percentage of people who have a certain disease die within a certain time after their disease was diagnosed? Ideally, we would like to use the date of disease onset as the beginning of the time period specified in the numerator in many chronic diseases, it may be difficult to determine precisely when the disease process began. we often use date of diagnosis as a surrogate measure for date of disease onset, because the exact date of diagnosis can generally be documented from available medical records. If the information is to be obtained from respondents, and the disease in question is a serious one, the date on which the diagnosis was given may well have been a life-changing date for the patient and not easily forgotten. the difference between case-fatality and a mortality rate 15     In a mortality rate, the denominator represents the entire population at risk of dying from the disease, including both those who have the disease and those who do not have the disease (but who are at risk of developing the disease). In case-fatality the denominator is limited to those who already have the disease. case-fatality is a measure of the severity of the disease. It can also be used to measure any benefits of a new therapy; as therapy improves, case-fatality would be expected to decline. 16    The numerator of case-fatality should ideally be restricted to deaths from that disease. it is not always easy to distinguish between deaths from that disease and deaths from other causes. For example, an alcoholic person may die in a car accident; however, the death may or may not be related to alcohol intake. An example 17     In a population of 100,000 persons, 20 have disease X. In 1 year, 18 people die from that disease. The mortality is very low (0.018%) because the disease is rare once a person has the disease, the chances of his or her dying (case fatality) are great (90%). PROPORTIONATE MORTALITY 18   proportionate mortality is not a rate of all deaths in a country, what proportion was caused by a certain disease 19 Deaths from heart disease as a percentage of deaths from all causes, by age group, United States, 2014 20 21 22    If we observe a change in proportionate mortality from a certain disease over time, the change may be due not to changes in mortality from that disease, but to changes in the mortality of some other disease. we may see that the proportionate mortality from cancer and from other causes has decreased in the population, but only because the proportionate mortality from other disease has increased. if all-cause mortality rates differ, cause-specific mortality rates can differ significantly, even when the proportionate mortality is the same. YEARS OF POTENTIAL LIFE LOST 23     Used for setting health priorities. a measure of premature mortality, or early death. YPLL recognizes that death occurring in a person at a younger age clearly involves a greater loss of future productive years than death occurring at an older age. Two steps are involved in this calculation for each cause, each deceased person’s age at death is subtracted from a predetermined (or “average”) age at death  the “years of potential life lost” for each individual are then added together to yield the total YPLL for the specific cause of death.  Years of potential life lost (YPLL) 24 Annualized years of potential life lost (YPLL) per 100,000 25 YPLL 26  YPLL can assist in three important public health functions:  establishing research and resource priorities  surveillance of temporal trends in premature mortality,  evaluating the effectiveness of program interventions WHY LOOK AT MORTALITY? 27      Mortality is clearly an index of the severity of a disease from both clinical and public health standpoints mortality can also be used as an index of the risk of disease mortality data are easier to obtain than incidence data for a given disease, and it therefore may be more feasible to use mortality data as a proxy indicator for incidence when a disease is mild and not fatal, mortality may not be a good index of incidence. A mortality rate is a good reflection of the incidence rate under two conditions:    when the case-fatality rate is high (as in untreated rabies), when the duration of disease (survival) is short. Under these conditions, mortality is a good measure of incidence, and thus a measure of the risk of disease. Age-adjusted death rates, by race 28 PROBLEMS WITH MORTALITY DATA: cause of death 29      Most of our information about deaths comes from death certificates. By international agreement, deaths are coded according to the underlying cause. The underlying cause of death is defined as “the disease or injury which initiated the train of morbid events leading directly or indirectly to death or the circumstances of the accident or violence which produced the fatal injury.” The underlying cause of death therefore “excludes information pertaining to the immediate cause of death, contributory causes and those causes that intervene between the underlying and immediate causes of death. total contribution of a given cause of death may not be reflected in the mortality data as generally reported; this may apply to a greater extent in some diseases than in others. PROBLEMS WITH MORTALITY DATA: quality of data 30     Countries and regions vary greatly in the quality of the data provided on their death certificates. Studies of validity of death certificates compared with hospital and autopsy records generally find higher validity for certain diseases, such as cancers, than for others. Deaths are coded according to the International Classification of Diseases (ICD) Because coding categories and regulations change from one revision to another, any study of time trends in mortality that spans more than one revision must examine the possibility that observed changes could be due entirely or in part to changes in the ICD. Other problems 31   Changes in the definition of disease can also have a significant effect on the number of cases of the disease that are reported or that are subsequently classified The denominator must be adjusted for those who are not at risk Comparing Mortality in Different Populations 32  An important use of mortality data is to compare two or more populations  one population in different time periods.     Such populations may differ with regard to many characteristics that affect mortality Age is the single most important predictor of mortality. Therefore methods have been developed for comparing mortality in such populations while effectively holding constant characteristics such as age. Adjusting for age 33   direct age adjustment. indirect age adjustment direct age adjustment. 34    a standard population is used in order to eliminate the effects of any differences in age between two or more populations being compared A hypothetical “standard” population is created to which we apply both the age-specific mortality rates from the 2 populations we eliminate any possibility that the observed differences could be a result of age differences in the population. age-adjusted rates 35     we derive the expected number of deaths that would have occurred had those rates been applied. We can then calculate the total number of deaths expected in the standard population for each population Dividing each of these two total expected numbers of deaths by the total standard population, we can calculate an expected mortality rate in the standard population if it had had the mortality experience of the 2 populations. Differences in age-composition of the population are no longer a factor. Problems 36      Although age-adjusted rates can be very useful in making comparisons, the first step in examining and analyzing comparative mortality data should always be to carefully examine the age-specific rates for any interesting differences or changes. These differences may be hidden by the age-adjusted rates, and may be lost if we proceed immediately to age adjustment without first examining the age-specific rates. Age-adjusted rates are hypothetical because they involve applying actual age-specific rates to a hypothetical standard population. They do not reflect the true mortality risk of a “real” population because the numerical value of an age-adjusted death rate depends on the standard population used. Selection of such a population is somewhat arbitrary because there is no “correct” standard population, but it is generally accepted that the “standard” should not be markedly different from the populations that are being compared INDIRECT AGE ADJUSTMENT (STANDARDIZED MORTALITY RATIOS) 37        often used when numbers of deaths for each age-specific stratum are not available. also used to study mortality in an occupationally exposed population the age-specific rates for a known population, such as all men of the same age, are applied to each age group in the population of interest. This will yield the number of deaths expected in each age group in the population of interest, if this population had had the mortality experience of the known population. for each age group, the number of deaths expected is calculated, and these numbers are totaled. The numbers of deaths that were actually observed in that population are also calculated and totaled. The ratio of the total number of deaths actually observed to the total number of deaths expected, if the population of interest had had the mortality experience of the known population, is then calculated. standardized mortality ratio 38 SMR 39