Ebola is obviously much less contagious than airborne viruses like influenza, and much more contagious than some other viruses, such as HIV, the virus which causes AIDS. What can we say at this point about how contagious Ebola really is?
The Official Statements
Through mid-September the WHO’s web page included this unfortunate paragraph, which is still widely quoted:
The risk of Ebola transmission is low. Becoming infected requires direct, physical contact with the bodily fluids (vomit, faeces, urine, blood, semen, etc.) of people who have been infected with or died from Ebola virus disease (EVD). [WHO 9/16/14]
As of September 20 this language had disappeared, and WHO’s current fact sheet has a more detailed and sobering description of how Ebola is transmitted between people:
Ebola … spreads through human-to-human transmission via direct contact (through broken skin or mucous membranes) with the blood, secretions, organs or other bodily fluids of infected people, and with surfaces and materials (e.g. bedding, clothing) contaminated with these fluids.
Health-care workers have frequently been infected while treating patients with suspected or confirmed EVD. This has occurred through close contact with patients when infection control precautions are not strictly practiced.
Burial ceremonies in which mourners have direct contact with the body of the deceased person can also play a role in the transmission of Ebola.
People remain infectious as long as their blood and body fluids, including semen and breast milk, contain the virus. Men who have recovered from the disease can still transmit the virus through their semen for up to 7 weeks after recovery from illness. [WHO 9/19/14]
The U.S. Centers for Disease Control (CDC) Transmission web page roughly parallels WHO’s fact sheet:
When an infection does occur in humans, the virus can be spread in several ways to others. Ebola is spread through direct contact (through broken skin or mucous membranes) with
Healthcare providers caring for Ebola patients and the family and friends in close contact with Ebola patients are at the highest risk of getting sick because they may come in contact with infected blood or body fluids of sick patients.
During outbreaks of Ebola, the disease can spread quickly within healthcare settings (such as a clinic or hospital). Exposure to Ebola can occur in healthcare settings where hospital staff are not wearing appropriate protective equipment, including masks, gowns, and gloves and eye protection. …
Once someone recovers from Ebola, they can no longer spread the virus. However, Ebola virus has been found in semen for up to 3 months. People who recover from Ebola are advised to abstain from sex or use condoms for 3 months. [CDC 9/20/14]
This post will try to tease out what these texts mean, and how well they square with the way in which the West African epidemic has developed. Public information needs to be clear, simple and not dangerously misleading. But it also tries to avert panic; it isn’t always the “whole truth.”
Reading Between the Lines
Both WHO and CDC state that contact with bodily fluids, etc. must be “(through broken skin or mucous membranes).” Apart from a cut this refers to the mouth, nose, eyes and anal and genital openings. What neither statement emphasizes is the importance of the hands as a means for the virus to get to a mucous membrane. A touch to the lips or the eye with a contaminated hand could do the trick, as could also a stray droplet of any sort of bodily fluid.
The examples given of contaminated surfaces differ oddly between the two descriptions. WHO mentions “bedding and clothing” while CDC mentions “needles and syringes.” Both obviously have the potential for transmission. Possibly the WHO fact sheet is directed to individuals while the CDC was thinking more about healthcare providers.
The CDC fact sheet says that “Ebola is not spread through the air.”
- It is obviously possible for someone to be infected by a droplet that passes “through the air” onto a mucous membrane. Various experiments have shown such transmission between animals in a laboratory setting. CDC acknowledged this in an earlier fact sheet, but explained, “While all Ebola virus species have displayed the ability to be spread through airborne particles (aerosols) under research conditions, this type of spread has not been documented among humans in a real world setting, such as a hospital or household.” [CDC 4/9/10]
- Ebola is obviously not spreading through the air anywhere near as easily as flu. This is the current phrasing of the CDC guidance for managing ill airplane passengers that originally was cited as evidence that the CDC believed in airborne transmission: “Ebola does NOT spread through the air like flu.” [CDC 9/20/14]
- The key question is whether Ebola is in fact being transmitted during the current epidemic through the air, and if so how significant this mode of transmission is, and how much it is affected by the condition of the patient. We’ll consider this further below.
- Mutation could in principle affect contagiousness by air, and we must remain vigilant about this possibility.
WHO advises that men use condoms for 7 weeks after recovery, while CDC recommends 3 months. That’s a pretty big difference! The Public Health Agency of Canada states that, “Ebolavirus has been isolated from semen 61 to 82 days after the onset of illness, and transmission through semen has occurred 7 weeks after clinical recovery.” [PHAC 9/20/14] This would seem to be a situation in which WHO is minimizing a risk that might exist for more than 7 weeks. At the moment sexual transmission by survivors is the least of our worries, but it may become important in the final phase of ending the epidemic.
The West African Experience
The current epidemic has provided a treasure trove of information about how Ebola does — and doesn’t — spread. Hopefully someone is collecting as much of this information as possible for scientific analysis. Only a few specific cases have been well enough described in the press for any conclusions to be drawn by outsiders, however. While anecdotal, they are nevertheless highly suggestive. For this purpose we are as interested in who didn’t get infected as in who did.
The Super-Spreader Funeral
All cases of Ebola in Sierra Leone have been traced to 14 women who were infected at the funeral of a traditional healer in Guinea. [NYT 8/28/14] It is customary at traditional funerals for people to wash, touch and even kiss the body, so it’s not surprising that they were infected. What’s interesting is that dozens of people attended the funeral — presumably quite a few more than the 40 who agreed to give blood samples — and it’s reasonable to suppose that many of them also had contact with the body. The corpse was clearly quite infectious, but even so it infected only perhaps 1/4 of those who attended the funeral.
The Nigerian Outbreak
Patrick Sawyer was a sociopath who caused the death of a dozen people, and very nearly unleashed Ebola on the most populous country in Africa. [EbolaStrategy 8/31/14] He was also a naturalized American citizen, and the Coordinator of the Economic Community Of West African States (ECOWAS) Unit of the Liberian Finance Ministry. [Note: Nigeria reported a total of 19 Ebola cases. I’ve been able to identify only 14 from the press. The others may be secondary infections plus a handful of additional primary infections, but they don’t much change the picture painted below. I’ll update this post if I’m able to sort out this discrepancy.]
The story begins with Sawyer’s sister, known as Princess, who was bleeding uncontrollably when her fiancé brought her to St. Joseph Catholic hospital in Monrovia. Her symptoms were recognized and hospital staff initially refused to touch her, but the Chief Administrator of the hospital, Brother Patrick Nshamdze, decided to treat her as if she presented a simple case of a miscarriage. [NationalChronicle 7/29/14] [FrontPageAfrica 8/13/14] Princess continued to bleed, however, and she was about to be moved to an isolation ward when Sawyer arrived. “He insisted that she be given a private room and plunked down $500 to secure it. He proceeded to personally change her gown and placed her in a wheelchair for the move. He was seen to get her blood on his own clothes as well as his shoes in the process.” [DailyBeast 8/14/14] Princess died on July 7. Understandably, Sawyer, Nshamdze and other healthcare staff who treated her without precautions became infected. Then the same thing happened when Nshamdze himself fell ill – doctors, nurses, a social worker and a lab technician fell ill. There’s no special information about contagiousness here, however, since there was lots of blood and few precautions.
Because of close contact with his sister, Sawyer was monitored daily and told not to leave Liberia. On July 20, after he began showing serious Ebola symptoms, including fever and vomiting, he got on a plane to Lagos, Nigeria, with a layover in Togo, as part of a nine-member ECOWAS delegation to a conference. He vomited repeatedly during the flight, then collapsed in Lagos airport. He was helped into a taxi and taken to First Consultant Hospital, where he initially told the medical staff that he had malaria and denied contact with any Ebola cases. Only two days later was he tested and found tentatively positive for Ebola, at which point he was quarantined.
One of the nurses who treated him, Obi Justina Ejelonu, became infected even though: “I never contacted his fluids. I checked his vitals, helped him with his food (he was too weak). … I basically touched where his hands touched and that’s the only contact. Not directly with his fluids. At a stage, he yanked off his infusion and we had blood everywhere on his bed … But the ward maids took care of that and changed his linens with great precaution.” Another hospital source “told [Front Page Africa] that in addition to yanking the infusion tubes, Sawyer took off his pants and urinated on the floor as nurses fled from his presence.” [FrontPageAfrica 8/11/14] He died five days later, on July 25.
A doctor who treated him, Dr. Ada Igonoh, also became infected. She gives an extremely detailed account of Sawyer’s stay at her hospital (and of her own illness and recovery) in a moving Front Page Africa article. [FPA 9/18/14] She felt that, “my contact with Sawyer was minimal. I only touched his I.V. fluid bag just that once without gloves. The only time I actually touched him was when I checked his pulse and confirmed him dead, and I wore double gloves and felt adequately protected.” The first several days in which Sawyer was treated for malaria without full precautions could explain her infection, however, as well as that of Dr. A.S. Adadeveoh, who treated Sawyer during the same period. [FPA 8/25/14]
“In total, Sawyer reportedly came in direct contact with 59 persons, 44 of whom were at the hospital… Sawyer came in contact with three ECOWAS officials – a driver, a liaison officer and a protocol officer. Also in the list are two nursing staff and five airport handlers.” [FrontPageAfrica 7/31/14]
Out of all his 59 contacts Sawyer appears to have directly infected just 5 people: The nurse and two doctors already mentioned, an ECOWAS official who picked Sawyer up at the airport and took him to the hospital, and an ECOWAS diplomat who was part of the group who greeted him on his arrival. All his other contacts did not get infected. That includes both passengers and crew on two flights, during which he was repeatedly vomiting, and everyone else in the Monrovia, Togo and Lagos airports. He was capable of infecting others, obviously, but on balance Sawyer was not very contagious.
The ECOWAS Diplomat
The ECOWAS diplomat escaped from quarantine in Lagos on July 26, traveled to Port Harcourt, and arranged to receive secret treatment in a hotel room by a local doctor, Iyke Enemuo. The diplomat recovered but the doctor was infected. There’s no indication that any other contacts of the diplomat were infected, however, so once again the healthcare providers bear the brunt of transmission. Apart from infecting Enemuo, the ECOWAS Diplomat was not very contagious.
Dr. Iyke Enemuo
Dr. Enemuo was yet another sociopath, following in Patrick Sawyer’s footsteps. He kept his contact with the infected ECOWAS diplomat secret and continued to pursue an active medical practice and social life, even after falling ill. WHO recounted the shocking details (apologies for the duplication from my first post):
After onset of symptoms, on 11 August, and until 13 August, the physician continued to treat patients at his private clinic, and operated on at least two. On 13 August, his symptoms worsened; he stayed at home and was hospitalized on 16 August. Prior to hospitalization, the physician had numerous contacts with the community, as relatives and friends visited his home to celebrate the birth of a baby.
Once hospitalized, he again had numerous contacts with the community, as members of his church visited to perform a healing ritual said to involve the laying on of hands. During his 6 day period of hospitalization, he was attended by the majority of the hospital’s health care staff.
On 21 August, he was taken to an ultrasound clinic, where 2 physicians performed an abdominal scan. He died the next day.
Nigeria had to track some 255 contacts, 60 of whom had high-risk or very high-risk exposure!
Of all those many contacts, Dr. Enemuo actually infected only 5 people: his wife (also a doctor), his younger sister, a doctor who treated him at the hospital, a pharmacy technician, and an older woman who shared his hospital room. [NigeriaTimes 9/19/14] [Reuters 9/1/14] He was certainly capable of infecting people, but considering the number of people he was in contact with after falling ill, Dr. Enemuo was not very contagious. Once again, caregivers and immediate family members were at greatest risk.
The Senegal Student
A 21-year-old college student was under observation in Guinea because he had helped care for several family members who had fallen ill with Ebola. Nevertheless, he traveled in a six-passenger vehicle to Dakar, Senegal, arriving on August 20. He stayed with his uncle in a crowded household for three days, then went to a local hospital, but did not mention his exposure to Ebola. He was initially sent home, but the following day “the Guinean health services reported ‘the disappearance of a person infected with Ebola who reportedly traveled to Senegal,’ according to Senegal’s health minister,” and the student was quarantined. [BBC 8/29/14] [Bloomberg 8/31/14] 67 close contacts were placed under observation: people he was in contact with during his journey, members of his uncle’s extended household, healthcare providers at the hospital, etc. [WHO 9/12/14]
The striking feature of this case is that none of the student’s 67 close contacts became infected! This is very different from other situations. My best guess is as follows: The student was never terribly sick, and has fully recovered. Consequently, he may not have had the same viral load as other patients, and may not have been throwing off bodily fluids as prolifically. It’s also possible that the main hospital in Dakar follows more rigorous universal precautions than those in the other affected countries. Whatever the reason, this is a striking instance of how non-contagious Ebola can sometimes be.
More than 240 healthcare providers have been infected during the West African epidemic, half of whom have died. [WHO 8/25/14] The great majority were local doctors and nurses who did not wear full protective gear, often because they were not yet aware that a patient had Ebola and sometimes because of a shortage of supplies. In a few cases, however, healthcare providers who were apparently following standard procedure nevertheless became infected. These cases raise concerns about the mode of transmission.
Dr. Kent Brantly
Brantly became infected working with patients at a Monrovia hospital operated by evangelical Christian organization Samantha’s Purse. He asserts that he followed all CDC recommended procedures, including wearing a standard N-95 face mask, the type also used by Médecins Sans Frontières/Doctors Without Borders (“MSF”). [MSF 7/28/14] This raises a question of whether Brantly was infected by airborne droplets that passed through his mask. This point is pressed in a tendentious but interesting article, written by a doctor who argues that Ebola healthcare providers should use the more expensive and uncomfortable P-100 (HEPA) face masks to fully protect against airborne droplets. [AmericanThinker 8/24/14]
The N-95 mask is designed to block 95% of “solid and water-based particulates (i.e., non-oil aerosols)”. The P-100 (HEPA) face mask blocks 99.97% of particulates and aerosols, whether or not oil-based. [CDC Jan, 1996]
- Maybe Brantly made some other mistake, of which he wasn’t aware. This seems frighteningly possible, with the odds of an error — small though they may be — relentlessly adding up as he worked day after day in close contact with Ebola patients.
- Maybe airborne transmission is indeed very rare, but Brantly got unlucky.
- Maybe airborne transmission is easier than we have thought, and MSF has been very, very lucky.
- Maybe the virus has mutated to become more easily transmitted by air — though obviously still vastly less easily than the flu.
The French MSF Volunteer
MSF announced on Sept. 17 that a French volunteer at one of its treatment centers in Monrovia had become infected, despite their strict infection-prevention protocol. She is the first international MSF staff member to be infected in the West African epidemic, although 7 local staff members had previously fallen ill, out of more than 2,000 MSF staff members in the region. The treatment center is turning away new patients pending an investigation into how the volunteer became infected. [Reuters 9/17/14]
From one perspective eight people out of 2,000 is less than half of one percent. But, as MSF President Joanne Liu had previously pointed out, the entire control effort can fall apart if healthcare providers feel personally unsafe.
We have no further information about how this volunteer may have been infected, but the announcement of the MSF investigation implies that it wasn’t due to anything obvious. Airborne transmission is a possibility.
The American Journalist
Added 10/6/14: An American journalist, Ashoka Mukpo, tested positive for Ebola on Oct. 2. He believes that he may have become infected when he was splashed while spray-washing a car in which an Ebola patient had died. [ABC 10/6/14] This seems credible and consistent. He is, by the way, being treated in the high-level biocontainment unit at the University of Nebraska Medical Center. This unit is the largest of four U.S. biocontainment units, and has 10 beds, which means that the U.S. has a total of less than 40 bio-containment beds in the entire country. Hmm… [NPR 10/6/14]
Ebola Transmission in the Developed World
At this writing there have been only two instances of transmission of Ebola outside of Africa, both involving nurses who cared for Ebola patients in the last stages of the illness. Both wore full anti-infection gear; this is not like the West African cases in which healthcare providers were infected before precautions started being taken, or because proper gear wasn’t available. Both nurses — at least initially — said they were not aware of any breach of the anti-infection protocol.
The Spanish Nurse
Added 10/6/14: A nurse (more precisely, a volunteer auxiliary nurse) contracted Ebola while treating a Spanish missionary in Madrid’s Carlos III hospital, where he died on Sept. 25. She is the first person known to have contracted Ebola outside of Africa. Reportedly she only entered the patient’s room twice, once after his death. [NYT 10/6/14]
We have no information about what precautions the nurse used, and whether a mistake may have been made. Update 10/8/14: The nurse initially said that she had followed the anti-infection protocol to the letter, which would have undermined confidence in the protocol and increased the anxiety of healthcare personnel. She then said that she “might have” touched her face with a contaminated glove as she was taking the protective suit off. [NYT 10/8/14] It’s reassuring — if also a bit too convenient — to have an explanation for how the nurse was contaminated.
The infection of the Spanish nurse is worrisome but not inconsistent with the experience in Africa, where hundreds of healthcare workers have contracted Ebola, and several Western medical personnel have been infected without noticing a failure of protocol. We do need to figure out why this is happening, but the extreme infectiousness of corpses and people in the last phase of the disease is known, and there’s nothing about this case which changes the picture.
The real issue in her case has nothing to do with contagion: it’s the fact that Spain took nearly a week to test her for Ebola after she reported that she was running a fever. This casual attitude will have to stop! Western countries have ample resources to contain Ebola, but it will bite anyone who treats it with laziness or contempt.
The Dallas Nurses
Added 10/14/14: Nina Pham, a Dallas nurse who gave extensive care to Thomas Eric Duncan throughout his hospitalization is the first person known to have contracted Ebola in the U.S. As noted above, she was not aware of any breach of protocol. This is of course worrisome, and affords further evidence that the CDC needs to focus more closely on the protocol, and the way in which training is given. But it isn’t a new issue. [NYT 10/13/14]
Again, the important issue isn’t contagion, but the fact that Ms. Pham wasn’t monitored as a contact. Fortunately, she self-monitored, detected a low-grade fever, and had herself admitted to the hospital. She tested positive for Ebola on the evening of 10/11.
Added 10/20/14: A second nurse who treated Duncan, Amber Joy Vinson, was confirmed as having contracted Ebola on Oct. 15. She had been self-monitoring and had reported a low grade fever, but was nevertheless allowed to fly to Cleveland on Oct. 10 and back to Dallas on Oct. 13. That was obviously stupid, and put at minor risk a large number of people in both cities and on both flights. But once again, the fact that an inadequately trained and protected nurse got infected from a later-stage Ebola patient is nothing new. [NYT 10/20/14]
Added 10/29/14: Both nurses have now been declared Ebola-free and discharged from hospital.
The Other Contacts of Thomas Eric Duncan
Added 10/8/14: It’s too early to say, but the fact that none of Duncan’s other contacts have yet shown symptoms is notable. He was very ill before being taken to the hospital the second time, and lived in close quarters with his girlfriend, her son, and her two nephews. He wasn’t as sick the first time he went to the hospital, but it seem clear that several doctors and nurses were in contact with him without taking exceptional precautions. It would be a remarkable confirmation of how infectious Ebola isn’t if all of these folks escape infection.
Added 10/20/14: Happily, nearly all of the 50 contacts with Duncan before his second hospitalization have now completed their 21-day quarantine period, with no Ebola symptoms! As well as being great news for them and for Dallas it confirms once again that Ebola is remarkably noncontagious until the final stage of the illness. Dozens Declared Free of Ebola Risk in Texas [NYT 10/20/14].
One nuance doesn’t change this conclusion, but could lead to a further little outbreak: As noted below, 5% of people infected with Ebola don’t show symptoms until more than 21 days from exposure. [NEJM 9/23/14] If someone released from quarantine turns up with symptoms after the 21-day period CDC might end up with egg on its face, and public confidence could be seriously eroded. I understand why 21 days has been chosen as a way of sending a simple message to the public but taking this approach is a somewhat risky strategy.
The Contacts of the Dallas Nurses
Added 10/29/14: Nina Pham apparently had only a few contacts, chief among them being her boyfriend. Amber Joy Vinson, however, flew to Cleveland and — after reporting a low-grade fever — back to Dallas. Vinson accordingly placed at theoretical risk a large number of people. It will be interesting to see whether any of these contacts became infected. My own guess — based on the other cases described above — that nobody else will come down with the disease based on contact with either of the nurses. If so, that should alleviate the public’s anxieties, but probably won’t.
The Bottom Line
Here are a few concluding observations:
- Even though the epidemic is growing exponentially, the individual cases we’ve been able to review support the official position that Ebola is not very contagious, and generally requires contact with bodily fluids or contaminated surfaces. Even a corpse only infected about a quarter of the attendees at a traditional funeral in which touching the body was customary.
- Caregivers and immediate family members are at highest risk, and others are at relatively low risk.
- Infectiousness depends on how ill the patient is, both from the perspective of viral load and the amount and character of emitted bodily fluids. There is no evidence of transmission before the patient becomes symptomatic; very little evidence of transmission before the last stage of the illness; but a high risk of transmission to caregivers and others who have direct contact with a patient during the last stage of illness.
- It isn’t clear to what extent the virus can be transmitted through the air, but:
- Casual contact such as sharing an airplane or even mingling in a social setting rarely or never seems to be enough to infect. Added 10/20/14: As noted above it’s astonishing that Duncan’s girlfriend and the three boys who also shared the apartment where Duncan stayed avoided infection.
- Something is infecting healthcare providers, even when they believe that they are following CDC guidelines. It is urgently necessary to determine why so many healthcare personnel who claim to be following CDC procedures are getting infected. We have to figure out what’s happening with that and fix it, whether or not it involves airborne transmission.
Update 10/3/14: A survey article by the WHO Ebola Response Team in the New England Journal of Medicine provides a lot more information, although it doesn’t greatly change our understanding of the situation. [NEJM 9/23/14] Here are a few key points:
- The mean incubation period is 11 days, and 95% had symptoms within 21 days of exposure, the recommended period for follow-up of contacts. Accordingly, 5% of infected contacts will first present symptoms after follow-up ends.
- The article estimates the current effective reproduction numbers at 1.81 in Guinea, 1.51 in Liberia and 1.38 in Sierra Leone. This basically means that each infected person is infecting on average between one and two more people, thus causing the epidemic to grow exponentially. See my earlier post, The Ebola Chain Reaction, for a more detailed explanation. The article rather cheerfully comments that : “This means that transmission has to be a little more than halved to achieve control of the epidemic and eventually to eliminate the virus from the human population. Considering the prospects for a novel Ebola vaccine, an immunization coverage exceeding 50% would have the same effect.”
- Without changes in control measures, their estimated doubling times range between 15 days for Guinea and 30 days for Sierra Leone.
- Case fatality has been 70% in Guinea, Liberia and Sierra Leone, in contrast with earlier reports closer to 55%. It was lower in Nigeria but the number of patients was so small that this might be a fluke.
Also, here’s a very clear statement, in the Oct. 2 New York Times, of what is considered “direct contact” with the bodily fluids of an Ebola patient: Understanding the Risks of Ebola, and What ‘Direct Contact’ Means
Update 10/8/14: An article in the LA Times collects questions about whether Ebola might be more contagious than is currently believed: [LA Times 10/7/14]
- Mutation is of course a wild card, which could make the virus somewhat or much more contagious. This is one of several good reasons for devoting resources to stopping the epidemic as soon as possible. But I don’t see any evidence of significant change in contagiousness, and nobody is claiming specific evidence of relevant mutations.
- What we do see is a pattern of poorly-explained infections of healthcare providers, especially when patients are in the last stage of illness. This might involve small highly-infections droplets passing through the air, and conceivably might lead to further tightening of the anti-infection protocol. Added 10/14/14: In any case one must welcome CDC’s plan to improve training and to send a team to any hospital treating an Ebola patient. [NYT 10/13/14]
- What we do not see is large numbers of infections of people who have casual contact with patients, or who have contact with patients before they have become symptomatic. So long as that pattern holds Western nations should have little difficulty keeping Ebola under control, so long as they treat it with the great respect and vigilance that it deserves.
Update 3/24/15: Liberia has identified a single Ebola patient who appears to have been infected by her partner, a male survivor. [NYT 3/24/15] This presumably came through his semen, which is known to contain Ebola virus for several months after recovery. This will make the disease a bit more difficult to eradicate, but is not cause for panic. The fact that just one infection of this type has come to light suggests that this mode of transmission is infrequent. Additionally, now that an example exists one might expect many “discordant couples” (in which the male is a survivor and the female is not) to use condoms, which should further reduce the risk. One question that calls for further research is how long the virus can persist in the semen of a male survivor, and whether a negative semen test is sufficient to make unprotected intercourse safe for the survivor’s partners.