For public health interventions, which one is more reliable Prevalence in absolute numbers or per population ( eg. Per 100000).
Error propagation
The reliability of a quotient depends on nummerator and denominator. The borders of the absolute error are the ratio of the Cartesian product of differences times moduli diveded by the square of the approximation of the denominator.
Therefore imprecise numbers of the population lower the error bounds.
For very rare events and a fluctuating population with unfaithful census of populaton the absolute number instead of a prevalence ratio should be preferred.
Prevalence by definition means number of cases as a proportion of total population.
However if you are following a fixed population over time, the denominator remains constant so use of absolute cases may have more impact from a public health media/ communications viewpoint
Reliability means the degree to which your measure represents what you are trying to measure so if the population denominator was changing eg through migration, public health impact on disease would be more reliably measured by the proportion rather than absolute numbers.
CANs are hard objects and BEs are not what they pretend to be.
Beg Your Pardon James. What figure is more trustable
- a strict correct absolute number (CAN)
or
- a ratio of a CAN and figure with broad error span (BES)?
The rules of error propagation power the aberration of the appraiser population with BES by square.
It's a common misapprehension, than ratios inform more accurate by comination of figures than these figures itself. The win of comparability is often payed by a loss of pungency.
Epidemiological studies do not name errors and age of the census of population
Since reliability is more of a factor of consistency, prevalence presented as proportion is more reliable than absolute value and use of proportion for comparison, if required, is much better than use of absolute values since denominators of different or even same population can vary.
Prevalence is the number of people who have the disease X. Ideally they should be very much the same.
Prevalence is defined by divide number of cases of the total population. Prevalence rate is more reliable represent magnitude of the problem than absolute number of cases
I agree completely with Ouma Simple.
Prevalence expressed in proportion is more reliable and more useful for comparisons.
The relative proportion by the total population is better that the alone figure found
The rough prevalence is influenced by the "Big Killers" and the "Big Cripplers". Therefore the prevalence doesn't reflect the pure risk of disease as incidence.
Standards and ratios of prevalence show pros and cons:
No standards: + real, simple - restricted comparability
age specific ratios: + comparable reasonable detailed information - many subtypes
age standardised ratios: + comparable total ratios - standards influence results, loss of information about subtypes
Most people have answered correctly, prevalence expressed in proportion will be reliable as it gives insight into the exact number having the disease out of the total population.
Thanks
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Probability
- Probability Theory
- Reliability