When I was little, I read a Reader's Digest book of great disasters, which included a segment on the Black Death. One of the things the book tried to do was explain, on a child's level, why it wasn't easy to figure out that rats and fleas were the source of the plague. You couldn't just look for patterns, because there seemed to be no pattern. Half a household might drop dead while the other half only got a little sick, or remained entirely healthy. Plague doctors who handled the sick every day lived another 20 years. The real spread of disease wasn't like the movies, where one person coughing means everyone in close proximity is doomed.
One reason for the emergence of strange non-patterns like this is something called "super spreaders"—basically, some people spread disease more effectively than others. The infamous Typhoid Mary is the poster child for super spreaders, but the effect has been well-documented in a range of infectious diseases and it goes beyond the simple story of one woman who infected thousands. In fact, what makes the super spreader phenomenon so fascinating is that it isn't an anomaly at all. Super spreaders are the primary way some diseases spread. The Contagions blog—which is all about the history of infectious disease—has a great post up about this.
Eventually new models arose like the “20/80″ rule that says that 20% of cases are responsible for 80% of the transmission and formed a core ‘high risk’ group. This model works well for some diseases but not all.
For pathogens that do rely on super-spreaders, the majority of cases will not transmit the infection to anyone. This can lead to a sense of false security because it seems poorly communicated. As Galvani and May assert, “heterogeneously infectious emerging disease will be less likely to generate an epidemic, but if sustained, the resulting epidemic is more likely to be explosive”. Super-spreaders tend to beget more super-spreaders, although most of the cases they generate will still not transmit the infection to anyone. For example, a super-spreader begets 30 cases, 3 (10%) of which become new super spreaders. The rest may transmit to 0-1 people.
Super-spreading has been documented for HIV, SARS (Sudden Acute Respiratory Syndrome), measles, malaria, smallpox and monkeypox, pneumonic plague, tuberculosis, Staphylococcus aureus, typhoid fever, and a variety bacterial sexually transmitted diseases.
And that brings us back to medical mysteries because, the Contagion blog explains, we don't know exactly why some people are super spreaders and others aren't—or why some people are more vulnerable to infection than others. So far, what we have to go on is a list of well-established correlations.