The current definition of public health surveillance is the ongoing, systematic collection, analysis, and interpretation of health data, essential to the planning, implementation and evaluation of public health practice, closely integrated with the dissemination of these data to those who need to know and linked to prevention and control. Now that's a pretty broad definition. For the purposes of understanding infectious disease epidemiology, the health data that we're most often concerned with are the prevalence or incidence of an infection. The prevalence of infection is the number of individuals in a population that are currently infected with a pathogen at any point in time. The incidence of infection is the number of individuals newly infected with a pathogen within a given period of time. Thus, for understanding the spread of infection within a population, prevalence tells us about how the total amount of pathogen out there that could be spread. While incidence tells us about the rate, or risk, a successful infection that occurs per unit time. In practice, both incidence and prevalence are often reported as a proportion of the total population. That is, the number of infections per 1,000 people, which allows direct comparison across populations of different sizes. In order to understand the dynamic process of epidemic spread, we really want to know the incidence and prevalence of infection. However, we are often only able to observe the incidence and prevalence of disease. That is, people that are currently expressing the outcomes of infection. This highlights an inherent challenge to the study of the dynamics of infectious diseases. We want to know how infection spreads, but more often than not, we are limited to our ability to see where disease is. For many infections the best measures available are the numbers of individuals seeking care in health clinics. This provides a measure of the incidence of infection, that is, the number of new infections in a period of time. Since these are necessarily only those individuals that develop symptoms of infection, this is very likely to be an under estimate of the true numbers infected. As an example, a 2009 analysis of the incidents of pandemic H1N1 influenza concluded that the 43,677 cases of H1N1 that were confirmed through laboratory tests in the US between April and July, likely represented 1.8 to 5.7 million cases of influenza. Largely because the vast majority of those infected did not seek care or have specimens collected. Though under reporting may seem to be a major challenge in the study of infectious diseases, it's relevance depends on the goals of a surveillance program. If the goal of the surveillance program is to estimate the total burden of infection, which is often the case when public health agencies need to set funding allocations for programs targeted at one disease over another, then failing to account for under reporting can have a significant impact on outcomes. However, if the goal of a surveillance program is to detect changes in the incidence of infection; say to evaluate whether new vaccines or other preventative measures are reducing infection rates, then under-reporting may not be limiting. In this case, the measure of program success is a reduction in the rate of new cases relative to past rates, rather than an absolute number. Measuring the numbers of infections that are reported as a consequence of individuals seeking care, that is, patients visiting doctors, is an example of a passive surveillance system. By passive here, I mean that the person or a group doing a surveillance is simply recording the information as it arrives. This contrasts with active surveillance, whereby the surveyor is actively going into the population of interest and recruiting individuals to assess whether or not they are infected. Passive surveillance is commonplace because it is comparatively easy and inexpensive. Countries around the world have lists of reportable diseases that get recorded into passive surveillance systems every time a patient is diagnosed. Thus, in principle, if all cases of infections seek care, then a passive system will identify every case. However, it is more likely the case that only some unknown fraction of individuals will seek care, and thus be recorded by a passive system. As a consequence, passive surveillance may be useful for monitoring relative changes in incidences of infection, but cannot directly inform the absolute numbers. Active surveillance, by comparison, requires that the surveyor select a subset, called a sample, of the population to evaluate and then quantify the numbers of infected, or newly infected individuals in that sample. Active surveillance, while more complicated to implement, has a number of important advantages. First, while passive surveillance most commonly records new cases of disease, that is a measure of incidents, active surveillance can be targeted to record either new infections or all current infections. That is either incidence or prevalence within a sample. Second, because the surveyor defines the sample to be measured, it is straight forward to scale these observations up to level of the full population. For example, if a simple random sample of five percent of the population is measured, and ten infected individuals are detected, then we would expect there to be about 200 infected individuals in the whole population. Third, active surveillance can be used to quantify infection directly, by enrolling individuals in the sample even if they are not expressing the symptoms of disease. This may be particularly important for infections that cause disease only rarely. For example, polio virus causes paralytic poliomyelitis, the paralytic disease of the central nervous system in less than one percent of infected individuals. Thus there may be quite a bit of infection in a population before a passive system would detect it. An active surveillance system that tested for infection in asymptomatic individuals however, might have a much better chance of detecting infection before it spreads too far. Again, the types of surveillance used to monitor infection and disease should be dictated by the objectives of a surveillance program. In practice, though each has strengths and weaknesses, there are roles to be played by both passive and active surveillance in the monitoring of the spread of infection and evaluating the impact of public health controls.
