by Katherine Wu
figures by Kristen Seim
We live in an increasingly wired world: with apps for every purpose imaginable, it has become easier and easier to share information and build global communities. In the wake of the recent Ebola pandemic, we have harnessed this technology to combat infectious disease, employing technological tools for diagnosis (as in the case of the parasitic disease loiasis) and epidemiological surveillance (Ebola virus and other hemorrhagic fevers). This new wave of technology has the potential to increase real-time data sharing and collaboration between scientists across fields.
—
“Outbreaks are inevitable. Pandemics are optional.” – Larry Brilliant, Skoll Foundation Global Threats Fund
With the help of modern technology, communication is now more efficient than ever. Phones have apps for just about everything; contacts are a mere scroll and tap away. As we take advantage of these virtual venues, it is now possible to streamline the distribution of information and tools in a sustainable, cost-effective manner. Even most avenues of healthcare have incorporated Internet databases and other forms of high-speed technology into their infrastructure. Technology has taken over every aspect of human lives – but can we harness its power to save them?
Counting worms: apps for rapid diagnostics
CellScope Loa, an iPhone app introduced by Dr. Daniel Fletcher’s group at UC Berkeley this past May, offers a novel way to rapidly diagnose the African eye worm Loa loa, the causative agent of loiasis [1]. This roundworm lives just beneath the skin in its human hosts, causing local inflammation and swelling as it moves. Both larvae and adult forms can reach high numbers as they replicate inside their human host, often passing visibly across the eye (earning them the name “African eye worm”); however, the infection, while unappealing, is not typically lethal. Rather, the biggest problems with Loa loa arise when it co-infects a human host with other, far more dangerous parasitic worms – typically those that cause lymphatic filariasis or onchocerciasis, two debilitating diseases that can cause severe disfigurement and blindness, respectively.
The drug of choice for lymphatic filariasis and onchocerciasis is Ivermectin, which kills the juvenile forms of the parasitic worms and clears them from the bloodstream. However, if the patient also happens to be infected with high numbers of Loa loa (>30,000/ml), Ivermectin can backfire – while it still kills the worms, the sudden death of thousands of these parasites in every organ in the body incites a strong immune response that can lead to hemorrhage, brain damage, and even death. Timely treatment for the latter two diseases, which affect 120 million and 37 million people worldwide, respectively, is essential to prevent permanent damage or lethality. Unfortunately, in many Central and West African communities, where populations of the parasites’ insect vectors[1] remain mostly uncontrolled, Loa loa co-occurs with lymphatic filariasis and onchocerciasis at rates over 20% [2]. Thus, it is essential for health workers to have access to rapid diagnostic tools for Loa loa when treating patients with lymphatic filariasis and onchocerciasis. The gold standard for confirming a Loa loa infection is extremely laborious: blood samples must be taken when the worm burden is highest in the patient, smeared onto a slide, stained with a dye to visualize the worms under a microscope, and painstakingly counted by a trained technician. This requires time and resources that are not reasonably available to these communities, especially in villages in which these diseases are endemic and highly prevalent.
CellScope Loa takes under two minutes and a few brief taps on the iPhone screen. Furthermore, based on preliminary trials on patients, CellScope Loa misdiagnoses fewer than 1 in 10 million cases, which makes it as good a diagnostic as smear staining, at a fraction of the cost and time [3]. Once a finger prick of blood is taken, the sample can be loaded onto the CellScope device, which mounts onto the outside of an iPhone. The app then uses the phone’s camera to record a magnified 5-second video and analyze the motion of parasites visualized in the blood (Figure 1). By tracking the shifting of individual blood cells as worms wriggle between them, the app can calculate the concentration of parasites in the blood, giving the user of the iPhone a clear, simple readout. Such data helps clinicians identify those patients who should seek treatment alternatives – for instance, opting for surgical removal of the parasites in lieu of oral Ivermectin.
Figure 1: CellScope Loa makes use of iPhone technology to rapidly diagnose Loa loa disease. The app uses the motion of individual red blood cells, disturbed by the passage of parasites between them, to sensitively estimate the number of worms.
The potential here is immense. While the technology is still in its infancy, once production reaches a larger scale, entire villages may someday have access to rapid, accurate screening and diagnostics, bypassing the need for data processing through microscopes in faraway labs. The tools deployed here could also easily be applied to other parasitic diseases that otherwise require extensive tissue analysis – tuberculosis and malaria, for instance, in which bacterial and parasite burden can factor heavily into diagnosis and treatment.
Similarly, scientists at the forefront of Ebola research are capitalizing upon the ability to pinpoint cases of an infectious disease in communities. Dr. Pardis Sabeti’s group at The Broad Institute of Harvard and MIT has developed Ebola Computational Assignment of Risk Estimates (Ebola CARE), another mobile app that can be deployed in the field by health care workers. Ebola CARE relies on machine learning to generate predictive models of mortality for Ebola virus: when doctors input symptoms and demographic information for patients, the app outputs a mortality risk estimation that can be used to inform subsequent triage and treatment [4].
When sharing saves lives
The utility of these mobile apps goes beyond individual diagnosis. Once the data is catalogued in a smartphone, data sharing is possible, and digital maps of disease cases can be produced to track their spread within a community. This much-needed technological tool was sorely lacking at the beginning of the most recent Ebola outbreak[2].
Ebola, spread only by direct contact with the bodily fluids of an infected individual, is a cruel disease – one that takes aggressive advantage of human compassion. The most common causes of transmission are caring for the sick and burying the dead, putting caregivers at risk and deterring health workers from entering afflicted villages. Even when health workers finally arrived in afflicted villages, they were chased out by fearful residents. The foreign officials’ demands were alarming to locals: stop washing the bodies of the dead, stop caring for the sick. The panicked atmosphere had bred misconceptions about how Ebola was spread – through curses, aerosolized disinfectants, bad spirits, and even through the relief workers themselves [5]. Resentment fueled silence, and without easily accessible, culturally sensitive means to track, monitor, and contain the disease, the consequences were lethal.
In an attempt to bridge the gaps in communication, social entrepreneur Camilla Hermann launched a mobile app called Assisted Contact Tracing (ACT) in late 2014. ACT allows people to self-report Ebola symptoms on even the simplest of cell phones, making it easier for health workers to monitor patients and their close contacts without having to travel to the affected area in person. Based on the symptoms, which are collected over the phone twice daily, health workers can ascertain the risk level of each individual in question and, importantly, track the virus’ spread. Furthermore, the initiative bridges the communication gap between medical professionals and susceptible populations, as phone calls and text conversations are initiated in the region’s native dialect. Such close surveillance of the disease’s time course is critical: with Ebola, there is a 21-day monitoring process (the length of the incubation period[3]), and a region can only tentatively be declared Ebola-free if no cases have been reported for double this time (42 days).
The effectiveness of this simple, sensitive approach has already been established: patient calls to health workers have been made a rate of 74 percent, especially high for protocols that rely on self-reporting [6]. And Hermann’s approach is feasible: the number of Africa’s mobile subscribers currently hovers around 1 billion [7], making cellular connection a particularly alluring option to monitor disease. In addition, information from patients and their close contacts is fed back to a local health care center through a modem that does not require Internet access, increasing accessibility. Developed in collaboration with Doctors Without Borders and recently endorsed by the National Ebola Response Incident Management System of the Republic of Liberia, ACT can now be established as a deployable tool for future outbreaks.
Although the Ebola outbreak is now considered mostly contained (Liberia was recently declared Ebola-free), this data rides on the heels of months of poor communication between health workers, researchers, and locals in countries in which Ebola remains endemic. In evaluating the response to the pandemic, officials reluctantly agree that Ebola would have been contained much faster and more effectively had there been appropriate surveillance mechanisms in place. As it was, however, information flow was stagnant at best. Though the outbreak likely stemmed from a case in December of 2013, the Center for Disease Control (CDC) was slow to acknowledge the outbreak, and case reporting was tenuous and sparse. By June of the following year, it was almost too late – Ebola was “out of control,” and their denial and lackluster response only paved the way for further transmission [5].
The real-time data crossroads
In the early months of the outbreak, glimmers of hope began to surface when Dr. Sabeti and her team of researchers began to gather and share data on the epidemic [8]. Specifically, Dr. Sabeti’s group sequenced the Ebola genome, taking blood samples from a handful of patients scattered across Sierra Leone. The genetic variation they observed led the group to reach several important conclusions: 1) New mutations were emerging in the virus; 2) human-to-human transmission, rather than repeated acquisition of the disease from animal sources, as in the case of the original infection, was responsible for its spread across West Africa; and 3) ongoing surveillance, genomic or otherwise, was imperative to contain the outbreak.
But the boldest and perhaps most important move Dr. Sabeti and her colleagues made in their research was outside of the confines of their groundbreaking publication in Science [9]. Dr. Sabeti’s team chose to publish their results in real time – as soon as the genomic data was generated, it was made available to other scientists across fields and disciplines, so others could access and utilize the data. Much of this data was released before the publication of their paper – a rare practice in the world of academia (Figure 2). Even a year later, Dr. Sabeti continues to be involved in relief efforts, collaborating with other scientists to implement tools to make it easier for clinicians in the field. In addition to the Ebola CARE mobile app, the group has begun an initiative to educate West African researchers about the genomics of infectious disease. For two years now, visiting scientists have trained with researchers at The Broad, learning lab techniques, molecular biology, and sequencing procedures. Ultimately, the team plans to introduce modern protocols to the medical communities in which viral hemorrhagic fever outbreaks are most relevant. By imparting knowledge to local health care workers, these and subsequent efforts may create a sustainable foundation for sequencing technology, such that economically neglected areas of the world will be better equipped for future outbreaks.
Figure 2: Despite mounting cases of Ebola, no new sequences were released between August 2 and November 9. [Graph adapted from http://www.nature.com/news/data-sharing-make-outbreak-research-open-access-1.16966]
In the battle against infectious disease, diagnosis and containment are often two of the highest hurdles to clear. The causative agents of infection have evolved to take advantage of the social nature of our species – spreading by direct contact, exchange of bodily fluids, proximity, interaction with pets and insects. We cannot always control how quickly pathogens spread; what is in our hands, however, is the effectiveness with which we communicate with each other. The recent innovations addressing the diagnosis, tracking, and containment of infectious diseases such as Loa loa and Ebola virus have blazed new trails for researchers and health care providers. In developing apps such as these, scientists are not simply taking advantage of modern technological tools – they are also helping to bridge gaps in culture and knowledge. Perhaps unexpectedly, by relaying information through phones and digital interfaces, clinicians are forging new and closer connections with their patients.
However, such trends can only continue if scientists are willing to open these lines of communication. Concerns about patient privacy and a lack of incentives to publish data in real time have hindered data sharing in advance of publication. As of June 2015, only 800 Ebola genomes have been published – about 3-4% of all cases [10]. But without breaking the barriers between different labs, investigators, and disciplines, technological advances will be limited in their impact. Rather than fearing the political implications of data sharing, we must realize that the answer to communicable diseases may just lie in communication.
Katherine Wu is a first year graduate student in the Biological & Biomedical Sciences Program at Harvard University.
[1] Flies in the case of loiasis and onchocerciasis, mosquitoes in the case of lymphatic filariasis
[2] For more reading on the recent outbreak, please see these articles: http://sitn.hms.harvard.edu/flash/special-edition-on-infectious-disease/2014/understanding-ebola-viral-mutations/and http://sitn.hms.harvard.edu/flash/special-edition-on-infectious-disease/2014/regulatory-approval-of-treatment-for-ebola-virus-a-u-s-and-european-perspective
[3] The period of time between infection by Ebola virus and the first manifestation of symptoms.
References
[1] Williams, SCP. “Microscope made from smart phone diagnoses deadly African parasite.” Science, May 2015. http://news.sciencemag.org/africa/2015/05/microscope-made-smart-phone-diagnoses-deadly-african-parasite (For the original study, please see http://www.ncbi.nlm.nih.gov/pubmed/25947164)
[2] “Mapping co-infection with Loa loa (RAPLOA).” WHO. http://www.who.int/apoc/cdti/raploa/en/
[3] Oaklander, M. “Your iPhone Can Now Tell If You Have a Parasite.” TIME, May 2015. http://time.com/3848427/worms-parasites-eyeworm/
[4] Dr. Andres Colubri, Sabeti Lab, personal communication.
[5] Sack, K et al. “How Ebola Roared Back.” The New York Times, December 2014. http://www.nytimes.com/2014/12/30/health/how-ebola-roared-back.html?_r=0
[6] Shallow, P. “An American woman’s fight to stop Ebola with technology.” CBS News, May 2015. http://www.cbsnews.com/news/an-american-womans-fight-to-stop-ebola-with-technology/
[7] Polwart, N. “Mobile Health Apps Have Role in Ebola Crisis.” InformationWeek Healthcare, August 2014. http://www.informationweek.com/healthcare/mobile-and-wireless/mobile-health-apps-have-role-in-ebola-crisis/a/d-id/1306617
[8] Preston, R. “The Ebola Wars.” The New Yorker, October 2014. http://www.newyorker.com/magazine/2014/10/27/ebola-wars
[9] Gire, S. et al. “Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak.” Science 345(6202): 1369-73 (2014).
[10] Dr. Nathan Yozwiak, Sabeti Lab, personal communication.
For more information about CellScope Loa, please see http://cellscope.berkeley.edu/.
For more information about ACT, please see http://www.odisi.org/.
Further reading about setting guidelines for data sharing in the research community:
Yozwiak, N. et al. “Data sharing: make outbreak research open access.” Nature, February 2015. http://www.nature.com/news/data-sharing-make-outbreak-research-open-access-1.16966
For further reading on the clinical data resources that have been made public regarding the Ebola outbreak: http://fathom.info/mirador/ebola/datarelease