CONTAINMENT AND SURVEILLANCE OF A HOSPITAL OUTBREAK OF EBOLA VIRUS DISEASE IN KINSHASA, ZAIRE, 1976.
MARGARETHA ISAACSON (1), J.F. RUPPOL (2), R. COLLAS (3), N. MATUNDU (4), TSHIBAMBA (4), K. OMOMBO (5)
1. The South African Institute for Medical Research, Department of Epidemiology, P.O.Box 1038, Johannesburg, South Africa 2000
2. Fometro, Kinshasa, Zaire
3. World Health Organization, Kinshasa, Zaire
4. City Hygiene Service, Kinshasa, Zaire
5. Department of Public Health, Kinshasa, Zaire
The clinical aspects of Ebola Virus Disease (EVD) in three nurses were described in a separate paper (1).
When the diagnosis of Marburg or Marburg-like disease was made, hospital acquired infection had already occurred in one nurse. At the time there were reports, later confirmed, of many people having become ill and died of this disease in the Equateur region of northern Zaire and there were then no known cases of recovery.
The health authorities were greatly concerned and a state of panic was imminent among sectors of the public.
The main thrust of the International Medical Commission's (IMC) efforts was therefore directed at containment of the disease in this city with a population of about 2 million people and the country's main international airport. These activities consisted on the one hand of the institution of containment measures aimed at preventing further hospital cross infection and on the other of a large scale surveillance programme of contacts and investigation of reported cases of suspect EVD.
From the moment of arrival of the two nuns and the priest at the Ngaliema Hospital, Kinshasa, Zaire, on 25 September 2976, some precautionary measures were taken to prevent spread of infection. Pavillion 5 in which they were accommodated has a number of single rooms with toilet and shower ensuite. All of these rooms surround an open courtyard which promotes good ventillation. Barrier nursing was introduced from the start and cotton gowns and cotton masks were worn when attending the patient. These were later replaced by disposable gowns and masks but as supplies were inadequate the gowns and the disposable plastic overshoes were hung up outside the door of the patient's room for re-use. It is noteworthy that Sister E.R. (Case 2) did not wear protective clothing when attending her patient. The disinfectant in use was 'Dettol' (chlorxylenol) which is known to have a narrow antibacterial spectrum and is unknown to have viricidal properties. When nurse M.N. was admitted on 13 October, quarantine was imposed on all the staff of Pavillion 5 but all other patients were either discharged or moved to other wards as they had not been directly exposed to the EVD cases. The staff was accommodated in Pavillion 5, together with a large number of contacts of the sick nurse. On the 16 October, the following additional precautions were introduced and existing ones modified to reduce further the risk of virus transmission :
1. Sufficient supplies of protective disposable clothing were procured to enable single use of items only. These included Balaclava type helmets which cover the whole head leaving only the face exposed. Full face respirators were obtained which provided airtight isolation of the exposed part of the face with air exchange through a filter canister which was at least 99,98% ef fective against 0,3 micron particulate matter at a flow rate of 85 liters per minute. The respirators were assigned to individuals and marked with their names. Staff members were urged not to share or exchange respirators under any circumstances in view of the potential danger of a face piece becoming internally contaminated with virus by a wearer during incubation period should he/she become infected. Goggles, to be worn with a disposable surgical mask, were issued to some staff members as there were insufficient respirators at first. Special attention was paid to the correct method of removing protective clothing after leaving the patient's room so that the potentially contami nated surfaces never touched the bare hands or face or uncontaminated mate rials and equipment. The gloved hands were first rinsed with sodium hypo chlorite, after which the respirator was removed and sponged off with sodium hypochlorite. The Balaclava helmet and overshoes were pulled off with gloved hands and the gown, which had cuffed sleeves, and the gloves were then peeled off in one single operation, Wearing a fresh pair of gloves, the gown was folded with the exposed surface inwards. All these items were discarded directly into a makeshift incinerator and burnt. The respirators were replaced in their containers which were also marked with the individuals' names.
2. The use of the hospital autoclaving and incineration facilities, at some distance from Pavillion 5, was rejected as this would have necessitated transport of virus-contaminated material with potential danger to other parts of the hospital. At first an open fire was used in the pavillion courtyard but this was soon replaced by an improvised incinerator constructed from a large oil drum. The drum was covered with wire mesh to prevent dissemination of the larger burnt particles. All disposable equipment and other waste, excluding excreta, were burnt immediately on leaving the patient's room, enclosed in plastic bags. A large galvanized iron bath was placed on bricks in the courtyard, a fire ]it underneath, and this was used to boil all non-disposable instruments, utensils and linen (Fig. 1). All these items were soaked in hypochlorite prior to boiling. The virus was assumed to have a sensitivity to heat similar to that of Marburg virus (2).
3. The patients used their own toilets until the terminal day or two of illness. Initially the toilets were flushed without special precautions. Later, it was recommended that use of the toilet was followed by pouring in half a liter of undiluted household bleach (sodium hypochlorite) and that flushing was to be delayed until the next time it was used. The concentration of the local product could not be ascertained but subsequently various products were obtained to enable a standard hypochlorite solution of known concentration to be made up. Excreta can be more effectively treated by autoclaving or closed incineration when the necessary facilities are available. When Case 3 could no longer leave her bed, bedpans were used and treated in a similar manner and then boiled.
4. 'Dettol' (chlorxylenol), in use in Pavillion 5, was replaced by sodium hypochlorite as a 1% or 4% solution (the latter in the presence of much organic matter) and by an iodophor disinfectant in a concentration yielding 450 ppm available iodine (3). These disinfectants were recommended in view of their known viricidal properties.
5. After the death of the first two patients their bodies were removed and the rooms locked up. Following the death of the third patient all three rooms were fumigated with formaldehyde vapour on four successive days. The rooms were entered with full protective clothing, the mattresses removed and burnt. Floors, walls and furniture, which were still blood-stained, were scrubbed and disinfected. Sphygmomanometers and stethoscopes were dismantled and the parts either boiled or soaked in a hypochlorite solution. Finally, all items were left in the sun for several days.
6. Disposal of bodies was carried out by wrapping them in cotton sheets impreg nated with a phenolic disinfectant. The fully wrapped body was sealed inside two large, heavy-duty plastic bags and then placed in a wooden coffin. Contact of relatives and friends with the body was strictly prohibited. The prepared body was however released for the funeral service and immediate burial. After the death of the last patient, the staff remained in voluntary isola tion in Pavillion 5 for a further 21 days. Morale, which was at a very low ebb when the IMC arrived, was rapidly restored by active participation of commission members in their activities in the pavillion, and by the pro vision of reading matter and other recreational materials. Some of the IMC members were close primary contacts of EVD and many were secondary contacts. Although it would not have been feasible for the IMC to carry out their activities while in strict isolation its members avoided un necessary contact with the general public and generally did not use public transport, nor visited restaurants or shops.
7. A negative pressure medical containment bed isolator was provided by the Canadian government. It was accompanied by a nurse specialist trained in assembling the unit and in its operation and the handling of patients in these isolators. The isolator was to be used in the event of a further case presenting in the Kinshasa chain of infection. Although no further cases occurred it served as an invaluable morale booster as hospital staff members could be assured of nursing a patient with little danger to themselves. Also, they would not be subjected to further quarantine. Several Zaire nurses and physicians as well as IMC members were trained in the use of the isolator.
8. As there was a certain degree of overcrowding in Pavillion 5, another pavillion was requested to be vacated in order to house all those contacts who were not needed for nursing or cleaning duties. These contacts, and those subsequently traced and brought in by the surveillance teams, were accommodated in Pavillion 2 in cohorts according to their last day of contact. Segregation between cohorts was strictly enforced in self-contained rooms with private bathrooms, toilets and verandahs, and in suites of rooms for the larger cohorts. Their size varied from 1 to 8 people. The contacts were given prophylactic malaria treatment and their temperatures were taken twice daily. A fever lasting for three consecutive readings was considered an indication for transfer to Pavillion 5 as a suspect case of EVD until proven otherwise. Although a few people developed pyrexias, these proved to be transient and transfers to Pavillion 5 were not necessary. Cohort isolation, commonly practised in neonatal nurseries, was adapted to this situation as it was believed that the occurrence of further cases among the primary contacts was almost inevitable. This would have imposed further hardships on the whole group if they were isolated as one large group with free intermingling. With cohorts formed according to date of last contact, it was anticipated that cohorts could be released in their entirety at the end of the 21 day period, enabling the preparation of their quarters to receive new contacts. Should one of the contacts in a cohort become ill, only the members of that cohort would require extension of the isolation period. Morale in Pavillion 2 was at first also very low, and aggravated by the boredom of the children. The provision of toys for the children improved matters. There was one pregnant woman in the group and she gave birth to a healthy infant while in quarantine. Mother and infant were separated from their cohort and isolated together for the remaining period.
Fig. 1 The courtyard of Pavillion 5 in which the AHF patients and exposed staff were isolated. Note the make-shift incinerator, boiler and nurses in full protective clothing including full face respirators.
By the time a diagnosis of Marburg-like disease had been made, it was no longer vital to trace the contacts of Cases 1 and 2 during their journey from Yambuku to the hospital in Kinshasa as the likely maximum incubation period had already elapsed without secondary cases having been reported.
Case 3, Nurse M.N. of the Ngaliema Hospital, had numerous local contacts and a surveillance program was established. The surveillance team comprised 4 medical practitioners and 4 health inspectors with drivers. Its function was to trace and classify contacts, bring in primary contacts to Pavillion 2 for quarantine, register and visit the secondary contacts, and check out reports of cases of suspicious illness. Initial problems posed by lack of transport and breakdown of telephone services were relieved by the provision of WHO vehicles and the establishment of radio communication. Primary contacts were defined as persons having had face to face contact with confirmed or suspect cases, or who had stayed in their house or eaten at the same table at any time during or since the 48 hours preceding onset of illness (4).
Secondary contacts were defined as persons having had a similar relationship to primary contacts. On 12 October Nurse M.N. visited the Foreign Affairs Ministry in order to finalize arrangements for an overseas study visit. She spent several hours waiting in the company of numerous unidentified strangers. The following day she became ill with severe headache and it is possible that she had been infective for one or more days. On 14 October she became feverish and went by taxi to the Mama Yemo Hospital which is the principal hospital of Kinshasa, serving a population of approximately 2 million people. Here she was examined, and blood was taken for malaria microscopy. She was then referred to the isolation hospital but was not admitted. Nurse M.N. proceeded, still by taxi, to the University Hospital where she was once more examined, and then sent home. The following day she was scheduled to go on duty but was instead admitted as a patient to her own ward in the Ngaliema Hospital where she died 5 days later.
Despite her presence in a crowded government department, and in equally crowded emergency rooms in several hospitals where examinations and laboratory tests were done without precautions, no secondary cases developed from these contacts. Among her close contacts were a 14 year old girl who ate from the same plate and a young man who shared a bottle of soft drink with her on the first day of symptoms. This provides further evidence for the belief that the Ebola virus is not highly infectious and requires very close contact, probably with blood or secretions, for its transmission.
Thirty seven primary contacts of Case 3 were traced and quarantined. The quarantine period was based on the then estimated maximum incubation period of 16 days as seen in Case 3. As few data were available at the time about the incubation period, it was decided to err on the safe side and prescribe a quarantine period of 21 days. From the 37 primary contacts of Case 3, a further 274 secondary contacts were traced to 44 addresses. A considerable number of both primary and secondary contacts could not be traced, either because their identity was unknown, or their address was unknown or incorrectly supplied. The concept of secondary contact was later discarded when it became clear from data collected in the epidemic area of the northern Equateur region that secondary contacts were at no risk of being infected unless and until they themselves became primary contacts.
The team investigated 15 false alarms which included cases of typhoid, amoebiasis, viral hepatitis, a fatal case of acute pulmonary oedema, and a case of carbon monoxide poisoning.
The preventive measures practised during this outbreak were dictated by our personal experience and that of others during previous outbreaks of actual or suspected outbreaks of Marburg disease, Lassa Fever and nosocomial infections in general (3,5,6,7).
It became evident in Kinshasa, and later in the northern Zaire epidemic area, that airborne dissemination of virus did not play a major role, if any, in the transmission of the disease.
It appears that the observation of the basic principles of aseptic technique or barrier nursing are probably effective in breaking the chain of crossinfection. However, this requires constant supervision and staff members who are well versed in these principles. It is emphasized that protective clothing nay itself become a dangerous source of infection if it is carelessly taken off and discarded. During the Kinshasa outbreak, therefore, meticulous attention was paid to safe disposal techniques. Formalin should not be used in an attempt to disinfect excreta as it has very poor penetration power and tends to coagulate the surface, sealing viable virus inside.
Of some concern was the low morale and despondency of the staff in quarantine after the death of Case 2 and the onset of illness in one of their colleagues, Case 3. The psychological reactions of both patients and staff during epidemics with high mortality when little can be done in the way of either prophylactic or curative therapy deserves some attention. A team spirit was therefore actively cultivated and the isolated hospital staff responded dramatically to the various measures to achieve this.
Active cognizance by the Zaire Minister of Health and the IMC was taken of their plight and the sacrifices they were willing to make, and they were kept it all times fully informed about IMC activities and-events related to the epidemic outside the hospital. Recreational needs also received constant attention.
Following the introduction of Ebola Virus Disease (EVD) into Kinshasa a local hospital-based chain of infection developed which ended with the infection of a local nurse at the Ngaliema Hospital.
Thirty seven primary contacts of the latter patient were isolated in cohorts according to the last day of contact, and no further cases of EVD developed. 'he surveillance teams which identified and traced these contacts also visited 174 secondary contacts and investigated a number of reports from Kinshasa and surrounding districts Of suspicious cases of illness or death, none of which were shown to be related to EVD.
Other containment measures included the use of disposable protective clothing together with respirators, viricidal disinfectants such as sodium hypochlorite and an iodophor, and careful disposal of excreta and other contaminated materials and equipment. Staff members involved in nursing and cleaning duties were instructed in the correct procedures to prevent auto- and cross infection.
1. Isaacson, M. et al. (1978) Clinical Aspects of Ebola Virus Disease at the Ngaliema Hospital, Kinshasa, Zaire, 1976. in Proc. Int. Coll. on Ebola Virus Disease and other Hemorrhagic Fevers (6-8 December, 1977, Antwerp). Ed., S.R. Pattyn. Elsevier/North Holland Biomedical Press B.V., Amsterdam.
2. Siegert, R. (1972) Marburg Virus. in Virology Monograph 11, Springer, Berlin.
3. Monath, T.P., Casals, J. (1975) Guidelines for the diagnosis and care of patients with Lassa Fever. WHO Vir./75.1.
4. Sureau, P. et al. (1978) Containment and surveillance of an epidemic of Ebola virus infection in Yambuku area, Zaire, 1976. in Proc. Int. Coll. on Ebola Virus Disease and other Hemorrhagic Fevers (6-8 December, 1977, Antwerp). Ed., S.R. Pattyn. Elsevier/North Holland Biomedical Press B.V., Amsterdam.
5. WHO (1974) Weekly Epid. Rec., 41, 341.
6. Gomperts, E.D. et al. (1978) An approach to the handling of highly infectious material in a routine clinical pathology laboratory and in a viral diagnostic unit. S. Afr. Med. J., 53, 243-248.
7. Clausen, L. et al. (1978) An approach to the hospital admission and clinical isolation of patients with dangerous infectious fevers. S. Afr. Med. J., 53, 238-242.
G.A. Eddy : In view of the low level of infectivity of Ebola virus, would you go through the rigorous containment practives, if you had to do this all over again ? Would you be as concerned as you obviously were, about spreading the infection to nursing and medical staff ?
M. Isaacson : This is a question that has engaged us on numerous occasions, and it is extremely difficult to answer. My personal opinion is that perhaps we ape overdoing things a little bit, on the other hand I also believe that we cannot afford doing less than the maximum precautions that are available. We cannot do it ethically, we cannot do it scientifically. So I would say that if I would be consulted again, I would indeed recommend the same procedures, although I personally might perhaps take certain liberties, I certainly would not recommend anything less than we have recommended during this outbreak.
S.R. Pattyn : I think we cannot afford to do less until we know more about virus excretion.
R.E. Shope I wonder whether there was any consideration given to the
possibility that arthropods perhaps mechanically, might transmit the infection
and if any arthropod precautions were taken.
M. Isaacson : Arthropod transmission was certainly considered but it was quite clear that, had this virus been arthropod transmitted, and knowing Zaire and its arthropod population, we would have had thousands of cases and not just a few hundred There was no evidence whatsoever for arthropod transmission. Therefore we did not take any specific precautions.
H. Bijkerk : Has there been any indication of a transmission by sexual contact, in view of the patients with positive seminal fluid for quite a long time ?
K.M. Johnson : Of the twenty people who survived this disease, I think 9 of them were males and they, in turn, during convalescence did not transmit the disease to any sexual partner. I also would like to make one other comment. With the total numbers that we heard about it must seem to much of this audience, a year afterwards, like an
G.A. Eddy : in view of the low level of infectivity of Ebola virus, would you go through the rigorous containment practices, if you had to do this all over again ? Would you be as concerned as you obviously were, about spreading the infection to nursing and medical staff ?
M. Isaacson : This is a question that has engaged us on numerous occasions, and it is extremely difficult to answer. My personal opinion is that perhaps we are overdoing things a little bit, on the other hand I also believe that we cannot afford doing less than the maximum precautions that are available. We cannot do it ethically, we cannot do it scientifically. So I would say that if I would be consulted again, I would indeed recommend the same procedures, although I personally might perhaps take certain liberties, I certainly would not recommend anything less than we have recommended during this outbreak.
S.R. Pattyn : I think we cannot afford to do less until we know more about virus excretion.
R.E. Shope : I wonder whether there was any consideration given to the possibility that arthropods perhaps mechanically, might transmit the infection and if any arthropod precautions were taken.
M. Isaacson : Arthropod transmission was certainly considered but it was quite clear that, had this virus been arthropod transmitted, and knowing Zaire and its arthropod population, we would have had thousands of cases and not ..just a few hundred. There was no evidence whatsoever for arthropod transmission. Therefore we did not take any specific precautions.
Bijkerk : Has there been any indication of a transmission by sexual contact, in view of the patients with positive seminal fluid for quite a long
. Johnson : Of the twenty people who survived this disease, I think 9 of them were males and they, in turn, during convalescence did not transmit the disease to any sexual partner. I also would like to make one other comment. With the total numbers that we heard about it must seem to much of this audience, a year afterwards, like an awful lot to do about maybe something that isn't so big. I don't think however that anyone who was there, could have at the time failed to appreciate how little was known and what the potentials were in both town and country and I cannot recall anyone think at the time that any of the measures, any of the efforts that were expanded to prevent the continuance of the chain in the capital city and to thoroughly document how in the rural areas this epidemic had ended, were superfluous. I don't think there was anybody in Zaire who felt that too much work had been done. In the Bumba zone for several months following the official government lifting of the quarantine, the commercial airlines still refused to fly, the people who ran the river boats between Kisangani and Kinshasa still wouldn't pull in to Bumba to take off cargo and the people of Zaire still retained a very great sense indeed of horror and anxiety about this whole happening.