Schistosomiasis — bilharzia, snail fever — is caused by a flatworm that lives in human blood vessels. People catch it by putting their skin in fresh water. Not by drinking it, not from another person: by wading, washing, fishing, swimming or collecting water from a pond, lake, canal or slow river where the right kind of snail lives.
It is the archetypal neglected tropical disease. A systematic review and meta-analysis of schistosomiasis and water resources development estimated the numbers at risk and infected on a global scale [1], with the burden overwhelmingly in sub-Saharan Africa [2] and falling hardest on school-age children. Its persistence is a poverty problem before it is a biological one: the people who get infected are the people with no alternative to the contaminated water [3].
It is also, for a disease this neglected, unusually elegant. The parasite's requirement that two worms find each other inside a human body creates a mathematical property — a transmission breakpoint — that makes elimination conceivable in a way it is not for most infections. That is the subject of the centerpiece.
Start here: the life cycle, because everything follows from it
The cycle has two hosts and two free-swimming larval stages, and every control measure attacks one of its steps.
An infected person passes eggs in urine or faeces. If those reach fresh water, each egg hatches into a miracidium, a swimming larva with hours to find and penetrate a specific freshwater snail. Inside the snail the parasite multiplies asexually, and weeks later the snail releases thousands of cercariae — the infective stage. A cercaria that meets human skin penetrates it directly, without a bite and without being swallowed.
Inside the person the worms mature and pair. Adult schistosomes are unusual among flatworms in having separate sexes: the male carries the slimmer female in a groove along his body, and the pair takes up residence in the veins draining either the bladder or the intestine, where they can live for years.
The damage is not done by the worms. It is done by the eggs. A female lays hundreds a day; some pass out to continue the cycle, and the rest lodge in tissue, where the immune response to them produces granulomas and, over years, fibrosis [4] [5].
Two species dominate, and where the eggs lodge is what separates them:
- Schistosoma haematobium — the urinary form. Eggs in the bladder wall cause haematuria, bladder wall thickening, obstructive uropathy, and in women genital lesions.
- Schistosoma mansoni — the intestinal and hepatic form. Eggs swept to the liver cause periportal fibrosis, portal hypertension, splenomegaly and variceal bleeding.
A third, S. japonicum, occurs in Asia and has a zoonotic reservoir that makes it harder to control; its genome revealed features of host-parasite interplay [6], as did the S. mansoni genome [7] and transcriptome [8].
The clinical picture has been reviewed repeatedly and consistently [9] [10] [11] [12]. What is worth emphasising is how much of the harm is chronic, cumulative and unattributed: morbidity has been quantified across sub-Saharan Africa [13], the disability cost of chronic helminth infection has been reassessed upward [14], and the impact of chronic schistosomiasis has been argued to be systematically unacknowledged — anaemia, undernutrition, impaired growth and cognition in children who are never diagnosed [15].
Pillar 1: measurement and diagnosis
Counting eggs, and why the count is the diagnosis
The standard method is to find and count eggs. For S. mansoni this means a Kato-Katz thick smear of stool, a technique published in 1972 as a simple quantitative device and still, half a century later, the field standard [16]. For S. haematobium it means filtering a urine sample and counting eggs on the filter.
The count matters as much as the presence, because schistosomiasis is a disease of intensity: a person with a handful of worms and a person with hundreds have different diseases, and programmes are targeted on community-level intensity.
The sensitivity problem is severe and it is worst exactly where it matters most. A modelling study of Kato-Katz sensitivity as a function of infection intensity estimated that at 100 eggs per gram of stool, a single smear detects about 50 percent of infections and two smears about 80 percent; at 300 eggs per gram, 62 percent and 90 percent [17]. Sensitivity for S. mansoni is dominated by missed light infections.
Field data make the same point brutally. Among 96 schoolchildren in a highly endemic area of Côte d'Ivoire, stool collected over five consecutive days with five readings per specimen gave a point prevalence of 42.7 percent on the first sample and a cumulative prevalence of 88.5 percent after maximum sampling effort [18]. The same study found that after praziquantel, day-to-day variation mattered more than before, because most remaining infections were very light and likely to be missed altogether by a single day's examination [18]. That is the diagnostic trap this disease sets: the moment a control programme succeeds is the moment its measuring instrument stops working. Day-to-day and intra-specimen variation have been analysed for helminth egg counts more generally [19] [20] [21].
Point-of-care alternatives, and what they actually deliver
A Cochrane review of 90 studies, 88 from African field settings, assessed the alternatives against microscopy [22].
For S. haematobium, the best performer was not an antigen test but a urine dipstick: detecting microhaematuria gave sensitivity 75 percent and specificity 87 percent across 74 studies and 102,447 participants, better than proteinuria (61 and 82 percent) or leukocyturia [22]. A strip that costs pennies and detects blood in urine remains the most practical large-scale screen for urinary schistosomiasis.
For S. mansoni, the point-of-care circulating cathodic antigen (CCA) test detects a parasite antigen in urine and performed at sensitivity 89 percent and specificity 55 percent against microscopy at a trace-positive threshold, across 15 studies and 6,091 participants [22]. That specificity looks alarming until the reviewers' own interpretation is read: the test misclassifies many microscopy-negatives as positive possibly because it is more sensitive than the reference standard it is being judged against [22]. When compared against a higher-quality reference using multiple samples, specificity rose. Antigen detection has a long pedigree [23] [24], lateral-flow formats now exist [25], and CCA has been compared directly with Kato-Katz in the field [26]; antibody-based diagnostics have also been evaluated [27]. For S. haematobium, however, the CCA test performed poorly — sensitivity 39 percent [22].
Measuring the damage rather than the parasite
Ultrasonography assesses the organ consequences directly: periportal fibrosis and portal hypertension in S. mansoni, bladder wall thickening and hydronephrosis in S. haematobium. Anaemia is a common and measurable consequence, and one of the harms most often left unattributed to the infection [15]; co-infection with malaria and other parasites is the norm rather than the exception in endemic areas [28].
Centerpiece: a simple simulatable model of the mating breakpoint
Most infections need a host. This one needs a host, a snail, and — because schistosomes have separate sexes — a mate.
That last requirement produces something unusual. Consider a population of humans in which worms are distributed with mean burden M. If worms are rare, a female worm may be the only worm in her host, and she can lay nothing. The parasite's reproduction is therefore not simply proportional to how many worms there are; it is throttled at low density by the difficulty of pairing. George Macdonald recognised this in 1965 and drew out its consequence — a transmission breakpoint [29] — and it became a central feature of the macroparasite framework Anderson and May built for helminth control by chemotherapy [30].
The model here derives the mating function rather than quoting it. Assume worms are distributed among hosts as a negative binomial with mean M and aggregation parameter k — helminth burdens are strongly overdispersed, a few people carrying most of the worms [31] [19] — that each worm is independently male or female with probability one half, and that a female reproduces if at least one male shares her host. Working through the generating function of the negative binomial gives the fraction of females that are mated:
φ(M, k) = 1 − (1 + M/2k)^−(k+1)
which is zero at zero burden, rises monotonically, and approaches one. The script that draws the figure checks this closed form against a direct Monte Carlo simulation of the same assumptions across 400,000 simulated hosts, so the algebra is verified rather than trusted.
The parasite's effective reproduction number is then its basic reproduction number discounted by that mating probability: R_eff(M) = R₀ φ(M, k).
Because φ starts at zero, R_eff is below one at low burden no matter how large R₀ is. There is always a burden below which the worm population cannot replace itself. That burden is the breakpoint, and it is an unstable equilibrium: above it the population recovers, below it the population goes extinct without any further intervention.
This is the teaching point, and it is genuinely important. For most infections, control means holding transmission down forever; for schistosomiasis, sustained mass treatment could in principle push the worm population below a threshold from which it does not return. That is the theoretical warrant for elimination as a goal rather than perpetual control [32].
The model also produces a result that is not obvious and is not encouraging. Sweeping the aggregation parameter shows that more aggregation lowers the breakpoint: clumping worms into fewer hosts helps them find each other, so the more overdispersed the infection, the further down a programme must push before the population collapses. Overdispersion — the very feature that makes mass treatment efficient, since treating everyone catches the heavily infected few — makes the endgame harder.
And the breakpoint falls steeply as R₀ rises. In the highest-transmission settings the target burden is very low indeed, which collides directly with the diagnostic finding above: the burdens a programme must drive the population below are precisely the burdens Kato-Katz is worst at seeing [17] [18]. A programme approaching its goal loses the ability to tell whether it has got there.
Three honest limits. The aggregation parameters and R₀ values are illustrative scenarios, not measurements for any particular setting. The model contains only one density dependence — mating — and so locates the lower, unstable equilibrium while saying nothing about the endemic level above it; real infections are also bounded by density-dependent worm fecundity and acquired immunity. And it is deterministic and spatially homogeneous, whereas the endgame of any elimination programme is a stochastic, patchy problem in which local extinction and reintroduction dominate.
Pillar 2: treatment and control
Praziquantel, and the fact that there is only one
Treatment is a single drug: praziquantel, usually given as a single oral dose. It is safe, cheap, and effective against adult worms of all the human species.
Reliance on one drug for a disease affecting hundreds of millions is a structural risk, and the field says so plainly. Praziquantel is the only drug used against human schistosomiasis on a large scale; the schistosome calcium channel is the only identified molecular target and the evidence for it remains indirect; there is no clinically relevant evidence of resistance to date, but worryingly low cure rates have been recorded in some African studies [33]. Susceptibility and resistance have been reviewed since [34], and the question of whether the strategy can rest on this drug indefinitely has been asked directly [35].
Two limitations matter clinically. Praziquantel is much less active against juvenile worms, so a dose given during ongoing exposure misses the immature parasites already in the body. And it does nothing to prevent reinfection — a treated child who returns to the same water is reinfected. Neither is a defect in the drug; both are reasons treatment alone is a control strategy rather than a cure for a community.
Mass drug administration
Because diagnosis is insensitive and the drug is safe and cheap, the strategy is preventive chemotherapy: treating whole at-risk populations, principally school-age children, on a schedule set by community prevalence rather than by individual diagnosis. This is the approach codified in WHO guidance on the prevention and control of schistosomiasis and soil-transmitted helminthiasis [36], delivered increasingly through integrated NTD packages that treat several diseases with co-administered drugs in a single round [37] [38] [39]. The contribution of mass drug administration to global health has been assessed on its own terms [40], and pre-school-age children — long excluded — have been addressed separately [41].
The snail, the water, and the latrine
Attacking the human end of the cycle alone leaves the other end intact. The complementary measures work on the environment.
Snail control was the mainstay before praziquantel and retains a role; the host-parasite relationship with the intermediate snail has been studied since the 1960s [42], snail distribution is a large-scale determinant of where infection occurs [43] [44], and climate change is expected to move it [45]. The most striking recent result is ecological rather than chemical: restoring a native river prawn that preys on the intermediate snail reduced human schistosomiasis transmission in Senegal [46]. Operational snail-control research has been organised deliberately [47].
Water, sanitation and hygiene address the two points where the cycle crosses between human and environment — eggs getting into water, and people getting into water. A review of the roles of water, sanitation and hygiene in reducing schistosomiasis found them consequential [48], consistent with the broader evidence that improved water supply and sanitation reduce helminth infection [49] [50] [51]. This is also the slowest and most expensive arm, which is why it is the one most often deferred.
Pillar 3: what is unresolved
Elimination versus control. The breakpoint says elimination is possible in principle. Whether the treatment intensity and coverage required are achievable in the highest-transmission settings, and sustainable long enough, is an operational question the mathematics does not answer [32] [52].
Diagnostics for the endgame. A programme approaching elimination needs a test that works at low intensity, and neither Kato-Katz nor urine filtration does [17] [18]. Antigen detection is more sensitive but its specificity against an imperfect reference is unresolved [22], and molecular methods are not yet field-deployable at scale.
Drug security. No resistance has been demonstrated clinically, but a one-drug strategy under sustained global selection pressure is exactly the circumstance in which resistance is expected. Monitoring cure rates in mass drug administration programmes and developing alternatives — artemisinins, and inhibitors of the schistosome-specific thioredoxin-glutathione reductase [53] — are both prudent [33] [34].
A vaccine. There is none. Candidates have been in development for decades, drawing on the observation that human immunity does develop with age and exposure — IgE responses correlate with resistance to reinfection with S. haematobium [54] — and vaccine efforts have been reviewed periodically [55] [56]. A vaccine would change the calculus entirely, because it would attack the parasite population continuously rather than in annual pulses.
The wider biology. Schistosomes are potent immunomodulators, and the same regulatory effects that let them persist for years appear to reduce allergic disease in infected children — an association reported for both S. haematobium [57] and geohelminths [58], and part of a large literature on helminth immunoregulation [59] [60] [61]. This is not a reason to be relaxed about infection, but it is a reason the biology is interesting well beyond the disease.
Dig deeper in lmmol
Schistosomiasis and malaria are the co-endemic pair: the same regions, the same rainy seasons, the same poverty, and frequent co-infection in the same children [28]. Both are vector-linked, though the schistosome's intermediate host is a snail rather than a biting insect, which is why control here attacks water bodies rather than air. For a third parasitic disease whose control has been driven by attacking the intermediate host, see Chagas disease; and for the shared arithmetic of transmission thresholds in a very different setting, see dengue and Lyme disease. The chronic anaemia and growth impairment of heavy infection connect to sickle cell disease as another cause of childhood anaemia in the same regions, and the hepatic fibrosis and portal hypertension of intestinal schistosomiasis make an instructive comparison with the non-infectious liver scarring in fatty liver disease. The full collection is at health.