The opinions expressed here are those of the author, Jim Kent, and do not necessarily reflect those of the University of California Santa Cruz or any of its units.
It’s been nearly a month since I wrote my first Ebola blog entry. Since then the world at large and myself in particular have learned more about Ebola. We have seen clearly that the virus can be transmitted within hospitals in developed countries. We’ve gotten more data showing that good hospital care including hydration, survivor plasma, and electrolyte balancing can save 75% of the patients, perhaps more if applied early. We’ve seen that, from a political point of view, it’s better for the Centers for Disease Control (CDC) to overreact than under-react. We see the epidemic continue to grow, but we also see some signs of its growth rate slowing, at least in Liberia. It seems a good time for a follow-up post.
At UCSC we’ll be adding new data and editing the Ebola Portal (http://genome.ucsc.edu/ebolaPortal/) in the coming weeks. Wikipedia has done such a great job synthesizing Ebola scientific knowledge that we’re dropping the Treatments and Vaccines section of the Ebola Portal in favor of a Wikipedia link. We’re continuing to encourage people to release Ebola viral and antibody sequences. We’ve added a new viral genome sequence from the smaller epidemic going on in the Democratic Republic of Congo (Maganga et al., 2014), and expect the first sequences from American patients soon.
In broader scientific terms, I think that most of the important medical, scientific, and epidemiological issues are now known. The challenge of how to formulate this knowledge into the most effective response is still a huge task. How can we minimize the loss of life with the resources at our disposal?
Many aspects of the epidemiology of Ebola are clear. In Africa as a whole, the time it takes to double the number of people who have been infected is about three weeks. In rural areas and affluent urban areas the doubling time is approximately four weeks, while in the shantytowns it is approximately two weeks. Epidemics in general follow an S-shaped curve, as shown in Figure 1 below. Initially there is a period of exponential growth. Approximately at the point where half of the people have become infected, the growth slows simply because there are fewer people left to infect. Even in the worst hit place in this epidemic, the shantytown of New Cru Town in Monrovia, the epidemic is still in the exponential growth phase on the left side of this curve. This is both good and bad: good because most of the people have not been subject to the certain pain and likely death of Ebola infection, but bad in that the epidemic will rapidly worsen.
Figure 1. A graph of the constrained growth equation that epidemics tend to follow in enclosed, freely mixing areas.
Within a single patient, the medical course of the disease is also relatively clear. After initial infection there is an incubation period of typically 9 days, which can be as short as three days and at least as long as three weeks before symptoms develop. The first symptoms are similar to those of many diseases – aches, fatigue, sometimes a headache, and sometimes stomach pains. After about three days of general malaise, usually a fever develops. The disease progresses rapidly in the next four days. During this phase there is intense diarrhea, usually vomiting, and sometimes bleeding. An adult patient will lose about 10 liters of fluid per day from these causes, if kept hydrated, and will often die from the effects of dehydration otherwise. After four days of intense symptoms patients will start improving if they are destined to recover, or deteriorate further if not. The recovery rates in Africa are only about 30%.
Taking care of an Ebola patient is a lot of work and is vastly complicated by the precautions caretakers must take to avoid becoming infected themselves. The patients are in considerable pain and subject to retching, spasms, and convulsions. For many patients, a madness sets in during the peak of the disease as well. Getting the patients to drink their 10 liters of electrolytes or stay attached to their IV lines, as well as clean up after them, is physically demanding and emotionally draining work. This is exacerbated by the need to wear a protective suit that gets so hot people can safely work in it for only 45 minutes without themselves getting dehydrated. In the U.S. hospitals, approximately 100 staff are required for a single Ebola patient. Doctors without Borders manages to get by with much fewer staff than this, but it is unrealistic to think that an Ebola patient can be managed with less than two staff per bed.
This is where we come to the fundamental conflict between the epidemiology and the medicine. Medically we want to treat every Ebola patient. The combination of hydration and plasma and/or antiviral treatment seems to raise the recovery rate from 30% to 75%, and is likely to improve further as our experience and tools for treatment grow. However, according to CDC estimates (corrected for under-reporting), as of 9/26/2014 there were 1500 people needing beds in Ebola treatment facilities in Liberia and Sierra Leone alone. We did not have the ~3000 support staff we needed then, and do not have the ~10,000 staff we would need for the ~5000 people estimated to need beds as I write this on Nov 3.
In medicine, generally prevention is far easier than treatment. For Ebola the most important prevention is keeping the patient away from other people during the most infectious phase when the patient is sickest, typically starting the day after the first sign of a fever and continuing until the patient dies or recovers. If the patient dies, the body is also exceedingly infectious. By and large the Africans have accepted the need to treat the body as hazardous and to bypass traditional funeral practices as a result. The big controversy in Africa right now concerns what to do with the patient during the infectious stage.
Ideally, patients would be brought into a treatment facility a day or two before they become highly infectious. This would have the dual benefit of isolating the population at large from infection and more than doubling the patient’s chance of survival. Unfortunately, because we don’t have enough people to treat patients this way, we have to pursue other courses of action as well that are not ideal for the people currently infected, but at least reduce the amount of people who will be infected in the future. Once we have vaccines in quantity, likely by March 2015, the situation will get much better. In the meantime though, to save lives, we have to consider a measure nobody really likes – quarantine.
Quarantine has become a bad word, in large part because most of the recent quarantines have been implemented so poorly. Quarantine is never going to be a joyful event, but if done carefully and with compassion, it need not be particularly unpleasant either. Certainly being quarantined is much more pleasant than catching Ebola or having friends and family die, and for the next several months at least, that is the alternative.
In general, people need food, water, and protection from extremes of temperature to live, and a degree of social contact with friends and family and a bit of entertainment to be happy. There is no reason that these can’t be provided inside of quarantine, and the cost of doing so is ever so much less than the cost of providing care for an Ebola patient.
The worst hit parts of Africa, and the ones in most need of quarantine, are the shantytowns. In a shantytown in the tropics, most structures are little more than a roof for shade and protection from the rain. Setting up structures such as these, capable of holding a family or social unit of about six with simple cots to sleep on, would not be hard and could be the basis of a quarantine unit. Food could be distributed in a central mess hall, and temperatures taken before one was allowed into the mess hall to eat. People showing fevers or other signs of sickness would be taken from the mess hall to a community care center where family could see patients. Ideally quarantine units of approximately 250 people could be set up in many places. The 250-person limit would reduce the spread of infection within a unit.
Once out of quarantine, ideally the dwellers of a shantytown would be moved into a refugee camp that would slowly grow to the size of the shantytown it is replacing. This camp would need a mess hall and a latrine system of some sort.
People would be invited, not forced, from the shantytown into the quarantine facility. If food, water, shelter, and minimal medical care are available, it is likely that the demand for going into such a quarantine facility would exceed the space available. A lottery would be a fair way to decide who gets in first.
After a certain point in time, everyone in the shantytown will either have passed through quarantine and into the refugee camp, have caught Ebola and either died or become non-infectious, proven naturally immune, or gotten very lucky. At this point the shantytown could be disinfected and the people from refugee camp could move back home. It seems likely that we may have a vaccine deployed as well by then.
Outside of the shantytowns, needed quarantines could be done in people’s own homes. In villages, a community care center coupled with contact tracing is all that is necessary. The traditional methods of contact tracing do work well outside of dense urban settings lacking basic infrastructure.
What would a community care center look like? The goal would be to have a place where the patients could, to the best of their ability, take care of themselves with limited help from survivors of Ebola and the bravest volunteers from their friends and family. The crucial parts of a facility are:
- Adequate stocks of oral rehydration fluids containing the correct balance of sugars, sodium, and potassium salts.
- “Cholera cots” (see Figure 2) that can efficiently and safely collect the patient liquid hazardous waste.
- A place to disinfect and dispose of the waste.
- Basic protection equipment and disinfection facilities for the workers.
- Water and simple food such as bananas and rice.
- Lamivudine or other mass-produced antivirals that don’t require refrigeration, if available.
- A fence so patients can’t exit until they’ve recovered and to keep out unprotected people.
Figure 2. A cholera cot – a must for treating diarrheal diseases in the tropics. (Image from Hesperian health guides.)
How well these community centers will work is perhaps the most uncertain part of this plan but, particularly with the cooperation of survivors, they may represent our best hope until vaccines are widely available. Socially they would need to be set up so that people could visit and talk through the fence to patients, but be located out of sight of the main habitations so as not to provoke despair. Community care centers have worked successfully in some Liberian towns, as described in detail in the Nov. 4, 2014 issue of Morbidity and Mortality Weekly Report (MMWR) from the CDC (Logan et al., 2014).
The CDC has done a lot of good work in containing this epidemic. Where they’ve faltered has been in portraying more certainty and perhaps more optimism than is warranted by what we know. Perhaps the CDC and leadership are worried that people won’t listen to them if they don’t convey absolute certainty; that if they don’t minimize statements of risk, people will panic. However, panic is normally a temporary condition. In the end, level heads that can reasonably appraise the situation will prevail. How can we appraise the situation, though, if we are not told the truth in all of its uncertainty and risk?
It is true that Ebola is mostly spread by contact with bodily fluids. It is true in previous, smaller epidemics that airborne spread between humans, if any, has played a minor role. However, it is wishful thinking, not science, to absolutely rule this out. With a disease as dangerous as Ebola, certainly it is better to err on the side of caution. Wearing a face mask on public transportation in an Ebola-infected area and washing one’s hands when one arrives back home or at work should be our advice, not — as Obama has said in videos aimed at West Africans — that you need not worry about catching Ebola on the bus if you live in an area where it is rampant. Wearing full body protection including a breathing apparatus should be the norm among Ebola medical personnel, and somewhat belatedly it has become so.
It is true that people with Ebola will mostly show a fever before the illness gets really serious, and vomiting and diarrhea start. However, the temperature increase one develops in response to an illness is highly variable across the population. Children in general spike higher fevers than adults. A noticeable fraction of adults, around 10%, don’t get fevers higher than 100 degrees even in the absence of medicine. A significant fraction of people are on anti-inflammatory medications for arthritis and other common conditions and don’t get fevers for this reason. In Africa, where presumably people tend to be less medicated than in the U.S., reports show that 11% to 13% of people sick enough with Ebola to take themselves to the hospital do not have a fever (Schieffelin JS et al., 2014; Who Ebola Response Team, 2014).
It is true that Ebola is mostly non-contagious before people reach the stage of illness where they show a fever (if one is going to develop a fever). Using a RT-PCR test, we can’t detect virus in the blood before the initial pre-fever symptoms of malaise, aches, fatigue etc. are felt. By the time fever shows, typically we do get solid RT-PCR results, but the viral levels measure only 10% of what they will the next day when the viral level typically peaks and the blood, at least, is maximally infectious (Towner et al., 2014). The viral loads in blood typically remain at the peak level for four days, and then either the patient dies, or the viral loads decrease and the patient recovers. If we assume (and it is an assumption) that a person’s level of contagiousness follows the blood viral load, then certainly most of the disease transmission occurs in the last four days, rather than the days leading up to and including the initial fever stage. Because there is a lag time before people notice that they have a fever and go to the hospital, how much of the transmission is likely to occur in the 8 hours after fever starts? Since the viral load will be rising from 10% to 100% over the course of the day, following an exponential progression, I’ll estimate the viral load on average during the first 8 hours after fever as 13% of peak, and the next 18 hours after fever as 33% of peak. With this I can estimate the viral load over time in the 8-hour window as:
Initial-transmission/(initial-transmission + later-transmission)
(8 hours * 13%) / (8 hours * 13% + 16 hours * 33% + 4*24 hours * 100%)
which comes to almost exactly 1%. So, while it is scientifically reasonable to estimate that 99% of the transmission will be avoided if people go into isolation relatively promptly after they’ve reached the stage of the disease usually associated with a fever, it is also reasonable to estimate that 1% of the transmission occurs before this stage. The clinical and epidemiological data suggest that it could not be much higher than this, but are not strong enough to say that it could be lower. Given the deadliness of the disease, it is prudent to consider people infectious at a low level even before the illness becomes severe.
If the world at large tended to under-react early in the course of this epidemic, for the most part this has changed. The CDC and others have tightened their recommendations and response in the USA. African nations and health organizations have been effective in keeping the spread of Ebola outside of Guinea, Liberia, and Sierra Leone to small, quickly extinguished outbreaks. The combination of popular education about how to avoid catching Ebola, contact tracing, and quarantine seems to be putting the brakes on the epidemic in the rural areas of West Africa. I do hope a system similar to the quarantine-into-refuge I describe here can be applied to the slums and shantytowns, and that these, together with community care centers, will help save many of those in even the hardest hit regions.
Logan G et al. Establishment of a Community Care Center for Isolation and Management of Ebola Patients — Bomi County, Liberia, October 2014. MMWR 2014;63(Early Release):1-3.
Maganga GD et al. Ebola Virus Disease in the Democratic Republic of Congo. N Engl J Med. 2014 Oct 15. [Epub ahead of print]
Schieffelin JS et al. Clinical Illness and Outcomes in Patients with Ebola in Sierra Leone. N Engl J Med. 2014 Oct 29. [Epub ahead of print]
Towner JS et al. Rapid diagnosis of Ebola hemorrhagic fever by reverse transcription-PCR in an outbreak setting and assessment of patient viral load as a predictor of outcome. J Virol. 2004 Apr;78(8):4330-41.
WHO Ebola Response Team. Ebola virus disease in West Africa–the first 9 months of the epidemic and forward projections. N Engl J Med. 2014 Oct 16;371(16):1481-95.