The epidemiology of hepatitis C, named as such in substitution of the former ‘non-A non-B hepatitis’, after the publication of the paper by Choo et al.  and subsequent clinical, epidemiological and laboratory analyses, has remained a formidable challenge, despite two decades of research. The recent papers by Caiaffa et al.  and Romano et al. , written from different perspectives and using different analytical methods, represent seminal contributions towards a better understanding of such a major public health problem, a condition that currently affects 130–170 million people world-wide . Caiaffa et al.'s paper addresses a key population in the spread of hepatitis C virus (HCV) infection—non-injecting drug users. A recent systematic review has shown that non-injecting drug users have HCV prevalence rates substantially higher than the general population . In the context of an optimal control of blood supplies in Brazil, Argentina and Uruguay since the late 1980s/early 1990s, and declining trends of both the number of people who inject illicit drugs and different infections/diseases such as HIV/AIDS and hepatitis C among them [6–8], the role of other at-risk populations in the dynamics of HCV must be assessed thoroughly by any public health initiative aiming to curb the spread of HCV.
The role of the parenteral route (i.e. the exposure of susceptible patients' blood to contaminated blood) is pivotal to the spread of hepatitis C and has been documented by a comprehensive literature in the field of epidemiology and molecular biology [2–4,8]. Notwithstanding, a substantial share of both chronic carriers and recently infected individuals have no history of parenteral exposure due either to blood transfusions (or organ transplants) or to the shared used of injection paraphernalia. The epidemiology of such additional infections remains obscure, and has been attributed to tattooing and piercing with non-sterile materials or to the shared use of straws among people who snort cocaine (in both cases due to the putative contact with small quantities of contaminated blood), and to unprotected sex (especially in the context of different co-infections). The precise role of each one of these factors remains a challenge for epidemiologists and molecular biologists. The paper by Caiaffa et al. provides consistent evidence about the role of shared straws in the spread of HCV among a pool of non-injecting drug users from Argentina and Uruguay, and also highlights the relevance of their interactions with injecting drug users and/or people living with HIV/AIDS, basically in the context of shared drug use .
In a different, but complementary, way the paper by Romano et al.  documents clearly that HCV has been spreading in São Paulo state, Brazil, following well-defined waves over time. Such successive waves correspond to the dissemination of different viral genotypes. After waves driven by blood transfusion in the 1980s/early 1990s a new wave, among which HCV-1a predominates, was found not to be linked to blood transfusion as were the former cases, but rather to sexual transmission. This new wave seems to follow a power law that makes evident a correlation of the infection rates and the size of the networks of the infected individuals . The paper by Romano et al. did not assess the shared use of straws, as did Caiaffa et al., so it remains to be explored whether the power law correlating infection rates and the size of infected individuals' networks is associated with unprotected sex, the shared use of straws by people belonging to these networks, or both.
The epidemiology of hepatitis C remains a deep puzzle yet to be elucidated fully. As well as its obvious complexity, the full understanding of HCV dynamics is a key challenge for both public health and internal medicine due to the relevance of the disease in different strata of the population all over the world and the seriousness of some of their long-term consequences, such as cirrhosis and liver cancer. Due to the undeniable role of social networks in the spread of any infectious diseases and the long-term evolution of HCV infection, such answers much probably will emerge from a combination of different methods, including classic epidemiology, ethnography, molecular biology and in-depth analysis of the dynamics of social networks over time.