Insidens dan Prevalens

01Core Definitions and Conceptual Framework

Incidence refers to the number of new cases of a disease that arise in a defined population during a defined time period. It is fundamentally a measure of risk: the probability that a healthy person will develop the condition. Because incidence only counts new events, the denominator must be limited to people at risk, meaning those who do not already have the disease at the start of the observation period.

Prevalence refers to the total number of existing cases (both newly diagnosed and previously diagnosed) of a disease present in a population at a given point in time or during a given period. It is a measure of disease burden. Unlike incidence, prevalence includes people who developed the disease years ago and still have it today. This means prevalence is influenced not only by how often the disease occurs, but also by how long it lasts.

The most commonly tested relationship between the two is captured by this equation:

$$P\approx I\times DP\; \backslash approx\; I\; \backslash times\; D$$

Where $PP$ is prevalence, $II$ is incidence, and $DD$ is the average duration of the disease. This equation holds true under steady-state conditions, meaning the incidence and duration are relatively stable over time and the population is not changing dramatically. Understanding this formula is essential because exam questions frequently test how changes in incidence or duration independently affect prevalence.

For example, if a new treatment keeps patients alive longer but does not cure the disease (increased $DD$) while the rate of new cases stays the same (stable $II$), prevalence increases. Conversely, if a new drug cures the disease rapidly (decreased $DD$) without affecting who gets it, prevalence decreases even though incidence has not changed.

02Types of Incidence

Cumulative incidence (also called incidence proportion or simply "risk") is the simpler version. You take a group of people who are disease-free at the start, follow them for a set period, and count how many develop the disease. The result is expressed as a proportion. For this measure to be valid, you must follow everyone for the entire period, meaning there should be no loss to follow-up. In practice, this works best for short observation windows or closed cohorts.

Incidence rate (also called incidence density or person-time rate) accounts for the reality that people contribute different amounts of follow-up time. A person who is followed for 3 years before developing the disease contributes 3 person-years, while someone lost to follow-up after 1 year contributes only 1 person-year. The denominator is the sum of all individual follow-up times. This is the measure used in most cohort studies and clinical trials, and it is the denominator concept tested when vignettes mention "person-years."

A critical distinction on exams: cumulative incidence is a proportion (ranging from 0 to 1), while incidence rate is a rate (with units of inverse time, such as per person-year, and can theoretically exceed 1 if the time unit is small enough).

03Types of Prevalence

Point prevalence is by far the most commonly tested form. When an exam question says "how many people currently have the disease" or "proportion of the population affected at this time," it is asking for point prevalence.

Period prevalence includes everyone who had the disease at any moment during a defined window. It is always equal to or higher than point prevalence for the same population because it captures people who had the disease at the start of the period, plus anyone who developed it during the period.

An important nuance: prevalence is not a rate. It is a proportion. It has no time dimension in its units. Test writers sometimes offer answer choices that conflate prevalence with incidence by including time-based units. If the answer is expressed as "cases per person-year," that is incidence, not prevalence.

04The Relationship Between Incidence and Prevalence

This section is arguably the most tested area for this topic. Understanding how prevalence changes when incidence or duration shifts is central to exam questions.

The classic exam trap involves a fatal disease like pancreatic cancer. Although the incidence may be meaningful, the prevalence is low because the duration from diagnosis to death is short. If a new drug extends survival from 6 months to 3 years without reducing incidence, the prevalence will rise even though the drug has not affected how many people get the disease.

Another commonly tested scenario: a disease like HIV/AIDS. Before the era of antiretroviral therapy, HIV had a relatively shorter duration (patients progressed to AIDS and died within a decade). With effective antiretroviral therapy, patients now live for decades. Incidence in many regions has remained stable or even declined, yet prevalence has risen because the duration of disease has increased dramatically.

05Practical Applications and Exam Relevance

A high-yield connection to other biostatistics concepts: the positive predictive value (PPV) of a test depends on prevalence. In a low-prevalence population, even a test with high sensitivity and high specificity will have a low PPV because most positive results will be false positives. This is why screening tests are most useful in populations with higher disease prevalence, and why confirmatory tests are needed when screening in the general population.

06Common Calculations on Exams

Exam vignettes typically present a small population with defined numbers and ask you to compute incidence or prevalence. Here is how to approach them:

Example 1: Cumulative Incidence
A town has 10,000 residents. On January 1, 500 already have Disease X. During the year, 200 new cases are diagnosed. What is the annual cumulative incidence?

$$\text{Cumulative Incidence}=\frac{200}{\mathrm{10,000}-500}=\frac{200}{\mathrm{9,500}}=0.021=2.1\mathrm{\%}\backslash text\{Cumulative\; Incidence\}\; =\; \backslash frac\{200\}\{10\{,\}000\; -\; 500\}\; =\; \backslash frac\{200\}\{9\{,\}500\}\; =\; 0.021\; =\; 2.1\backslash \%$$

Note: the 500 existing cases are subtracted from the denominator because they are not at risk of developing the disease (they already have it).

Example 2: Point Prevalence
Using the same town: on December 31, how many people have Disease X? Assume 50 of the original 500 cases died during the year and 30 were cured.

Existing cases at end of year = (500 - 50 - 30) + 200 = 620

$$\text{Point Prevalence (Dec 31)}=\frac{620}{\mathrm{10,000}}=6.2\mathrm{\%}\backslash text\{Point\; Prevalence\; (Dec\; 31)\}\; =\; \backslash frac\{620\}\{10\{,\}000\}\; =\; 6.2\backslash \%$$

Note: the denominator for prevalence is the total population, not the population at risk.

Example 3: Incidence Rate with Person-Time
Three participants are followed for different durations: Person A develops disease at year 2 (contributes 2 person-years). Person B is lost to follow-up at year 3 (contributes 3 person-years, no disease). Person C completes the 5-year study without disease (contributes 5 person-years).

$$\text{Incidence Rate}=\frac{1\text{ case}}{2+3+5}=\frac{1}{10\text{ person-years}}=0.1\text{ per person-year}\backslash text\{Incidence\; Rate\}\; =\; \backslash frac\{1\; \backslash text\{\; case\}\}\{2\; +\; 3\; +\; 5\}\; =\; \backslash frac\{1\}\{10\; \backslash text\{\; person-years\}\}\; =\; 0.1\; \backslash text\{\; per\; person-year\}$$

The critical detail tested here: once a person develops the disease or is lost to follow-up, they stop contributing person-time to the denominator from that point forward.

07Traps and High-Yield Pearls

The most common way students get questions on incidence and prevalence wrong is by confusing what belongs in the denominator. For incidence, the denominator must exclude people who already have the disease at the start of the observation period, because they are not "at risk" of developing what they already have. For prevalence, the denominator is the entire population, including those with and without the disease.

The second major trap involves the $P\approx I\times DP\; \backslash approx\; I\; \backslash times\; D$ relationship. Exam questions will describe a clinical scenario (a new drug, a new vaccine, a change in survival) and ask what happens to prevalence. Students who do not think through whether the intervention affects incidence, duration, or both will select the wrong answer. A treatment that prolongs life increases prevalence. A cure that shortens disease duration decreases prevalence. A vaccine that prevents new cases decreases incidence, which then decreases prevalence.

The third trap is failing to recognize that prevalence is not useful for studying causation. Because prevalence mixes old and new cases, it cannot distinguish between factors that cause the disease and factors that simply prolong it. A factor associated with high prevalence could be a risk factor for developing the disease, or it could be a factor that makes the disease last longer. Only incidence can separate these.

Finally, students sometimes confuse incidence rate with cumulative incidence. If the question mentions "person-years" in the denominator or describes participants entering and leaving the study at different times, the answer requires an incidence rate calculation. If the question describes a closed group followed for a fixed period, it is asking for cumulative incidence.