The language of infection and disease

An epidemic is rarely produced by a single cause. Field investigation usually finds three forces acting together: a fresh supply of susceptible people, an agent that has become better at spreading, and a lapse in the measures that were holding the disease down. The first force is the susceptible population, and measles is the clearest illustration. Its basic reproduction number — the average number of people one case infects in a population with no immunity — is roughly twelve to eighteen, so a community needs about 92 to 95 per cent immunity before transmission stops, a threshold calculated as 1 − 1/R₀. When vaccination coverage slips below that line, as happened in several countries after 2015, the susceptible pool refills within a few birth cohorts and outbreaks return although the virus itself has not changed. Migration of non-immune people into an endemic area, waning immunity after years without exposure, and immunosuppression caused by HIV achieve the same result. The second force is the agent. Influenza demonstrates antigenic drift, the slow accumulation of mutations in its surface proteins that lets the virus escape antibodies from previous seasons, and antigenic shift, the reassortment of gene segments between human and animal influenza viruses that produced the 2009 H1N1 pandemic. Cholera shows another mechanism: the genes for its toxin are carried by a bacteriophage, so an otherwise harmless environmental Vibrio can acquire virulence in a single infection. The third force is a breakdown in control — conflict and displacement, a damaged water system as in Haiti in 2010, insecticide resistance in mosquitoes, a failure of the vaccine cold chain, or a drug shortage. Nursing acts on all three: the vaccination clinic refills the susceptible pool in reverse, notification and line lists detect the change first, and hand hygiene, isolation and safe water hold the line until the response catches up.

Most figures in an outbreak report are built from two counts. The attack rate is the proportion of people exposed to a source who become ill: cases among the exposed divided by the number exposed, multiplied by one hundred and expressed as a percentage. The denominator is the exposed group, never the whole population. In the funeral meal above, 54 of the 120 people who ate became ill, an attack rate of 45 per cent. Dividing the same 54 cases by the district’s 18,000 inhabitants would give 0.3 per cent, a figure that describes the district’s background experience and says nothing about the meal or about the risk run by the people who ate it. The rule has three practical consequences. First, an attack rate needs a defined exposure and a defined period rather than a time denominator, because the exposure itself fixes the period. Second, when spread is person to person the same arithmetic is applied to susceptible contacts and called the secondary attack rate: if 12 of 20 household contacts who were not already immune develop the disease, the secondary attack rate is 60 per cent, and that figure measures how efficiently the organism travels inside a home. Third, the denominator must be corrected by removing people who were already immune or were not truly exposed. Attack rate is easily confused with two neighbouring measures. Incidence rate counts new cases in a population at risk over a stated period, usually per 100,000 per year; prevalence counts all existing cases at one moment. Case fatality rate is a different proportion again — deaths divided by diagnosed cases. In the same outbreak, 2 deaths among 54 cases give a case fatality rate of about 3.7 per cent, and that figure is the population-level shadow of virulence, just as the attack rate is the population-level shadow of infectiousness. For the nurse the calculation is not an abstraction: a food-specific attack rate, say 70 per cent among those who ate the salad against 12 per cent among those who did not, is what identifies the vehicle, and a line list with person, place and time is what makes it possible.

The reservoir is the natural habitat in which an organism normally lives and multiplies and on which it depends, and it must be distinguished from the source — the particular thing from which this patient was infected — and from the vehicle, the water, food or needle that carried the organism across. Cholera and malaria share the same vocabulary of reservoir, portal and route, yet they demand opposite action lists, and the reason is the reservoir. Vibrio cholerae, serogroups O1 and O139, has a mainly human reservoir of cases and short-term carriers, with an environmental component in brackish water where the organism attaches to copepods — tiny aquatic crustaceans — and plankton. Its portal of exit is the faeces, its route is faecal–oral through contaminated water, food or hands, and its portal of entry is the mouth. Because the reservoir is human and the route is water, control is water and sanitation: chlorination with a free residual of about 0.2 to 0.5 mg per litre, protected wells, latrines sited well away from the water source, and hand hygiene with soap, supported by oral rehydration salts and antibiotic treatment of severe cases. The patient is also a source, so faeces and soiled linen must be handled as infectious waste. Malaria works differently. Plasmodium falciparum, P. malariae, P. ovale and P. vivax have a human reservoir of gametocyte carriers — people whose blood carries the sexual stages the mosquito takes up — and P. knowlesi is zoonotic, meaning it is maintained in animals such as macaques and can spill over into people. The parasite must pass through the female Anopheles mosquito, which bites mainly between dusk and dawn and in which the sexual cycle is completed; the mosquito is therefore an essential part of the chain, not a passive vehicle. Control acts at three points at once: against the vector, with insecticide-treated nets, indoor residual spraying and larval source management; against the parasite in the human reservoir, with artemisinin-based combination therapy (a drug combination whose core is a fast-acting artemisinin derivative), chemoprophylaxis (preventive drugs taken before or during exposure) and intermittent preventive treatment in pregnancy (timed preventive doses given to pregnant women); and at the moment of the bite, with repellents, protective clothing and screens. Draining a breeding site is a barrier that no medicine can place, and no insecticide net will stop cholera. When the reservoir is an animal, as in rabies, brucellosis, plague or yellow fever, treating human cases never removes the infection, so measures must reach the animal or its vector — dog vaccination for rabies, rodent and flea control for plague. When the reservoir is the environment, as with Clostridium tetani in soil or Legionella in water systems, the answer is engineering, disinfection or immunisation rather than isolation. Reading a public-health report therefore begins with one question: what is the reservoir?