Week 4: Assignment 2

PubH 6003

“AN OUTBREAK OF HEMORRHAGIC

FEVER IN AFRICA”
(this assignment is based on a case study provided by the CDC EIS and modified by M Ulfers for MPH@GW course 6003)

Please read each problem carefully and show your work –especially if you are hoping for
any partial credit! Clarity in your wording is a must. Avoid using jargon unless you are
completely sure you are doing so correctly .
All work should be neat and legible.
Round each of your answers to the nearest 10th decimal point (e.g., 3.258 = 3.3).

1
 AN OUTBREAK OF HEMORRHAGIC FEVER IN AFRICA

OBJECTIVES
After completing this case study, the student should be able to:

1. List the criteria for deciding whether a field investigation is warranted;

2. Describe the key tasks involved in investigating an epidemic of unknown cause, and, given a
real outbreak, perform each step;
3. Describe the components of a case definition, and develop a case definition for a new
disease;
4. Given the appropriate data, perform descriptive epidemiology and explain the importance of
this step.

THERE ARE A TOTAL OF 68 POINTS FOR THIS CASE STUDY

Note: This exercise is designed to “step” you through this outbreak investigation. Therefore
answer each question given only the information presented up to that point.

The Bumba Zone is in the middle Zaire river basin, an area consisting predominantly of tropical
rain forest. About 275,000 persons live there, mostly in small villages with fewer than 500
persons. The people are avid hunters and come in contact with a variety of wild animals.
Dysentery, malaria, filariasis, measles, amoebiasis, pneumonia, tuberculosis and goiter are some
of the common endemic diseases to this area.
Since the first of September 1976, 17 patients at the Yambuku Mission Hospital in the Yandongi
collectivity (county) and one Belgian missionary midwife employed by the hospital had
developed an illness characterized by fever, vomiting, abdominal pain, and bloody diarrhea that
rapidly progressed to death. The officer reported that the illness appeared to be spreading among
the remaining 16 hospital staff members, as well as persons living along the roads leading from
Yambuku.

Question 1a(1 pt.): What is the first question that needs to be answered before even
considering an investigation? Are the numbers real?

Are we prepared to safely do field work? (Do we have the resources and information needed
before we begin?)

On the 19th of September, 1976, the chief medical officer of the Bumba zone in northern Zaire
(now the Democratic Republic of the Congo) radioed the Minister of Health in Kinshasa to
report that an exceptionally lethal disease had become epidemic.

Question 1b(1pt): What other factors justify launching an investigation in this scenario?

The type of symptoms and the high infection rates among those within the hospital and the
fact that it has become and outbreak.
Exponentially lethal and spreading.

2
Question 2(2 pts): After having decided to investigate, but before considering an exact
course of action for the investigation, what other information might be useful to have?
Name at least 2 things…

The means of the spread- is it direct contact, vector-bourne, water-bourne, etc?

What type of infectious disease is this? (Viral, Bacterial, etc.)
Is this an existing disease?
Establishing the number of people infected and establish contact tracing to see who may have
been exposed.
Lab test? Treatments? Describe what cases look like? Are cases related by type of vehicle?
Training of filed staff?

Question 3 (4 pts.): Before departing, list and briefly describe (a sentence each) the last 4
accepted steps of an outbreak investigation.
After describing the cases by person, place, and time, the next step is to develop the hypothesis
and evaluate that hypothesis against established facts. Next, plan and conduct a case-by-case
analytic study. Implement control and prevention efforts among the involved communities. The
final step is to prepare and disseminate a report to be published and distributed by the appropriate
organizations.

23 September - 3 October 1976

Epidemiologic teams, composed of Ministry epidemiologists and microbiologists and medical

officers of two international medical missions to Zaire, were sent to the region. Thirty-two more
persons with the disease had been hospitalized at the Yambuku mission hospital and these
patients were examined. The illness was characterized by high temperature (> 39°C.), headache,
bloody vomiting and diarrhea, chest and abdominal pain, arthritis, and prostration leading to
death, usually in three days time. Jaundice was not present. Liver biopsy and blood specimens
were collected from the patients and were sent to the World Health Organization (WHO)
reference laboratories for further evaluation. On September 30th, the Yambuku hospital was
closed because 11 of the 17 staff members had died of the illness. On October 3rd, Bumba Zone
was quarantined.

Question 4a (2 pts.): At this time, what broad categories of causative agents would
you be considering? Name only 2.
The causative agents to be considered include viruses and parasites.

Question 4b (1pt): Aside from the lethality, what epidemiological data is particularly
striking so far?
Aside from the lethality, the short incubation time, the extremely high incidence rate, and high
infectivity rates as shown throughout the hospital patients and staff. With symptoms similar to
Cholera, this illness spreads quickly and progresses rapidly into a debilitating disease in the
individual.

3
October 1976
Among infectious agents, an enteric bacterial illness or a viral hemorrhagic fever fit best with the clinical
symptoms. Specific viral agents might include enterovirus, flavi-and other arboviruses, arenavirus (Lassa
fever), Congo-Crimean fever, or an unclassified agent. (you are being given this information as you
would not be expected to know it as a new epidemiologist!) It is also beyond the scope of this case study
to discuss specimen collection and laboratory methods for making a diagnosis. For your information,
when one may be dealing with an unknown infectious agent, specimens from multiple sites should be
considered (e.g., blood, CSF, respiratory tract, Gl tract [stool]), as well as tissues from fatal cases (e.g.,
liver, lung, brain, etc.).

A previously unclassified virus, similar to the Marburg virus that causes a type of hemorrhagic
fever, was isolated from liver specimens sent to WHO reference laboratories. A serologic test
was quickly developed.

Question 5 (3 pts.): What modes of transmission of illness should be considered in

light of this information? List only 3.

Direct contact methods include transmission through person-to-person touching or vector-bourne

through an arthropod bite. This illness may also be spreading through unclean drinking water.

13-20 October 1976

Following contact with a Belgian missionary nurse who had been transported to the capital for
medical care, a Zairian nurse in Kinshasa became ill and died. The Zaire Government asked for
the formation of an international commission to assist in the investigation of the disease. On
October 19th, 1976, following the first meeting of these consultants, a survey team was sent to
Bumba Zone. In addition to cases from whom virus had been isolated or those with positive
serology, the consultants identified other individuals with illnesses ranging from death following
an illness of headache, fever, abdominal pain, vomiting, and bleeding, to only headache and
fever following contact with another ill person.

Question 6a (4 pts): Given this spectrum of disease, how would you define a case
for this investigation? (i.e., develop a case definition)

An illness caused by a virus and characterized by fever, abdominal pain, vomiting, headache, and
bleeding. This must include all individuals, regardless of demographic factors, located in the
Bumba zone in northern Zaire that were exposed or felt symptoms of this virus as early as
September 1, 1976 and ending before October 25,1976. This includes people who only had a few
of these symptoms and came in contact with an ill person but may not have a positive serology.
This excludes non-infected people or people who do not live in this region. Laboratory criteria
includes positive confirmation via blood sample. Time- go back 1 month before first case.

4
The investigators chose to divide cases into three types:

• A proven case was a person from whom virus was isolated or demonstrated by electron
microscopy or who had an indirect fluorescent antibody titre of at least 1:64 to virus within
three weeks after onset of symptoms.
• A probable case was a person living in the epidemic area who died after one or more days with
two or more of the signs and symptoms .
• A possible case was a person with headache and/or fever for at least 24 hours, with or without
other signs and symptoms, who had contact with a proven or probable case in the preceding
three weeks.

Question 6b (1pt): Particularly considering the case definition for a probable case, what
can you expect the case fatality to be?

Based on the case definition and characterization of symptoms, we expect the case fatality to be
high.

Question 7 (2 pts.): How would you proceed with the investigation at this point? (Describe
the next step)

The next steps after resolving a case definition include identifying and accurately counting cases
to evaluate by person, place, and time. This would include all infected individuals in the Bumba
zone during September to October 1976.

5
The teams conducted active case finding in over 250 villages in the region of Yambuku. As a
result, they identified 463 cases meeting one of the case definitions. Of these cases, 38 met the
definition for a proven case, 280 for a probable case and 155 for a possible case. A summary of
the proven and probable cases by date of onset of disease is given below in Table 1:

6
Question 8 (6 pts.): Graph an epidemic curve. Indicate those cases which were fatal and
those which survived. (either use Excel or draw on a separate sheet of graph paper) Be sure
and label your x and y axes. Always use a histogram in drawing your epi curve. Make sure
that you choose the appropriate units of time for your x axis. (in the live session you may
want to just sketch a rough one to discuss)

Question 9a (3 pts.): Describe the epidemic curve in terms of the information it gives
about the outbreak.

The number of new cases per day does not reach above twenty-five, making the whole of the
outbreak a very short time period. When it peaks on day twenty-four, it rapidly declines
thereafter. Knowing that this disease is spread via direct contact means that we can label this as
either a common source or propagated outbreak.

Time period, Peak, Case fatality rate over time, mode of transmission difficult to deduce from
curve, What kind of outbreak is this?

7
Question 9b (1 pt.): What does it suggest about the nature of the disease (i.e. possible types
of outbreak as indicated by the shape of the curve)?
Because this may be a propagated curve with a short lifespan, this disease spreads rapidly from
person-to person and has a high infectivity rate. The irregular spikes and secondary cases that
lead to the peak may indicate the different outbreaks of the virus in different locations (i.e. other
villages). Say what it is not and make a case for what you think it is! Person-to-Person

The Yambuku Mission Hospital was established by Belgian missionaries in 1935 in the Yandongi
collectivity (county), one of seven in the Bumba zone. It serves as the principal source of
medical care for 60,000 people in Yandongi and adjacent collectivities. Because it maintained a
good supply of medicines, people passing through Bumba zone frequently traveled long
distances to attend clinics there. At the time of the current epidemic, the hospital had a staff of
17, including a Zairian medical assistant and three Belgian nurse midwives. There was an active
prenatal clinic that treated 6,000-12,000 patients each month.

Question 10 (4 pts.) : As mentioned earlier in the exercise, the Yambuku hospital had
closed on September 30 following the death of 13 of the 17 staff members. Mark the date of
hospital closing with an arrow on your epidemic curve and calculate 2 case-fatality rates:
First for through September 29, then for September 30 onward. Discuss plausible
explanations for the changes in rates from the information you have.

Case fatality rate through 9/29/1976: 245 total deaths/252 total infected X 100 = 97.2% case
deaths

Case fatality rate from 9/30/1976: 38 total deaths/66 total infected = 57.6% case deaths

From 9/30 onward, case numbers were declining and there was a smaller pool of infected
individuals. Because the hospital was closed, sick people stayed home in isolation rather than out
into the vulnerable population. Increased Immunity, Change in the way individuals spread the
disease, etc.

The following table lists the number of proven and probable cases according to sex and age:

Table 2: Distribution of Hemorrhagic Fever by Age and Sex, Zaire, 1976

Male Female Total
Age
Number Percent Number Percent Number Percent
(years
<1 10 3% 14 5% 24 8%
1 - 14 18 6% 25 8% 43 14%
15-29 33 10% 60 19% 93 29%
30-49 57 18% 52 16% 109 34%
50 + 23 7% 26 8% 49 15%

8
 Total 141 44% 177 56% 318 100%

Question 11a (2 pts.): Carefully consider the above data. What, if any, conclusions can you
draw from this table?

We do not have enough information from this table alone to draw decisive conclusions. However
from this data it seems that the virus disproportionately infects individuals ages 15-49 with
females at slightly higher rates than males across all ages excluding ages 30-49.

Question 11b (2 pts): What can you NOT conclude from this table alone?

This table only displays possible and probable causes and does not give definitive mortality or
survival rates.
Don’t know the risk for certain age groups, cannot make comparisons.

Question 12 (2 pts.): What additional information, if any, would help you to analyze these
data and be able to draw more conclusions?

The attack rates, mortality rates, and specific time periods during data collection
would help draw a better conclusion.
Use population census.

Table 3: Population in the Epidemic Zone By Age and Sex, Zaire, 1976

Age
(years) Male Female Total
<1 800 850 1,650
1-14 8,200 8,150 16,350
15-29 5,500 6,000 11,500
30-49 6,250 6,750 13,000
50 + 3,000 4,500 7,500
Total 23,750 26,250 50,000

Question 13 (9 pts.): Using the data in Tables 2 and 3, calculate age- and sex- specific
attack rates (per 1000) for each group in Table 4. What conclusions
can you draw from this information? (hint: you may want to
compare some risks mathematically)

9
 Table 4: Hemorrhagic Fever Attack Rates (Per 1,000) By Age and Sex, Zaire, 1976
Age (years) Male Female Total
(per 1000) (per 1000) (per 1000)
<1 14.5
1 - 14 2.2 2.6
15-29 6.0 8.1
30 - 49 9.1 8.4
50 + 7.7 6.5
Attack rate for total 5.9 6.7 6.4
pop.

The attached figure shows the location of villages that had one or more cases during the
epidemic, with the overall attack rate (in per cent) listed for each. These villages contained
42,264 (85%) of the estimated 50,000 persons living in the epidemic zone. The remaining 7,736
inhabitants live in eight non-infected villages. Other than Kinshasa, no towns outside the
epidemic area were found to contain cases.

10
 Figure 1.

Question 14 (2 pts.): What conclusions regarding place of onset and risk of disease by
place seem warranted?

The closer the individuals were to the point of outbreak, the more likely they were to contract
and spread the virus as evidenced by the non-infected villages outside of the endemic zone.

Question 15a (6 pts.): Summarize the information from all data given up to this point
regarding person, place and time and risk of disease. Be thorough but concise!

We are gathering information on an outbreak in the Bumba zone in northern Zaire of a virus
characterized by lethal symptoms such as fever, abdominal pain, vomiting, headache, and

11
bleeding. This outbreak period spans September 1,1976 to October 25, 1976. So far we have seen
an epidemic curve evidencing short incubation time, high infectivity rate, and very high
mortality rates among persons infected. Anyone who comes into contact with an infected person
is at high risk of contracting the virus.

Question 15b (2pts): Describe the virus in terms of suspected pathogenicity and virulence
from what is known so far.

This virus has suspected high rates of both pathogenicity- in its ability to produce visible,
diagnoseable illness levels in its host, and virulence- the quantitative measure of the virus to
cause severe disease. These can be recognized by the severity of symptoms and the lab tests
results detailing how much the virus can replicate inside the host.

Question 16a (2pts): One reason that the epidemic didn’t spread further is in general
victims did not spread the disease to many other people. How do you
explain this given what you know of the disease?

With the symptoms presented and course of the disease within the human body indicate a short
infectious period. The hosts typically die from the disease before it is able to reach a high enough
dose to infect the next individual.
R is a function of the Ro and immune coverage—we don’t expect any immunity here BIT Ro is a
function of duration.

Question 16b (2pts): What response probably ended the epidemic and why?

Quarantining all infected individuals and isolating the area to prevent visitors/non-infected
people was the response that probably stopped the spread, ending the epidemic. When the virus
no longer had viable healthy hosts, either by killing the infected host or creating an immune
response, it ceased to spread.

The following table presents the distribution of incubation periods among the 318 cases.

Table 6: Distribution of Incubation Periods*,

Hemorrhagic Fever, Zaire, 1976

12
 Incubation
Period # Cases
(Days)
3 5
4 12
5 18
6 58
7 22
8 24
9 85
10 80
11 9
12-15 5

* Interval from first contact at hospital and/or

person with disease to date of onset

Question 17 (4 pts.): Using the data in Table 6 describe, and interpret the data on
incubation period of the disease. (In reality, drawing really precise
conclusions from this may require more medical knowledge – which as an
investigator on a team you would be able to seek out!)
From the table we can infer that the incubation period is extremely short, only a matter of days
between case spikes. From the time someone is infected to the time they begin showing
symptoms looks like it is about or less than a week. The doubling rate looks like it may be as
short as a few days. The curve suggests a bimodal- 2 peaks- distribution which could be
explained by differences in the age of the cases at various points.

EXTRA CREDIT (2 pts): Very briefly discuss why this “original” outbreak of Ebola –
when there was so much less knowledge about it- stayed fairly contained in contrast to the
most recent Ebola epidemic West Africa.

There are several factors at play involving the spread of the Ebola virus. During the original
outbreak, the villages were small making the virus was easier to contain within those
populations. Air travel was not as prevalent or readily available as it is now mitigating the spread
because there was simply nowhere to go for missionaries or native residents. The Ebola virus has
most likely mutated since then possibly increasing the lethality or infectious quality.

13
Emphasize that one must perform additional analysis when one is confronted with a bimodal
curve. Graphing the data by time or site of acquisitin would be particularly helpful here.

14
CONCLUSION

The investigators conducted a case-control study within the hospital to identify factors associated
with spread of infection. All 85 cases who had contact with the Yambuku mission hospital but
not with another case had received one or more injections at the outpatient service or the general
medical wards. Fewer than 1% of controls had received injections at the hospital during the
epidemic.

The index case occurred in a 44-year-old male instructor at the mission school who presented to
the outpatient clinic at Yambuku Mission Hospital on August 26th with a febrile illness thought
to be malaria. This man had recently returned from a tour of the Mobaye-Bongo zone in the
northern equator region. He was given chloroquine by parenteral injection on 26 August. His
symptoms worsened on September 1st, and he died on September 8th.
The next 9 cases which occurred during the first week of September were all among individuals
who received injections at the outpatient clinic of the hospital. In total, 22 of the females in the
15-29 year old group acquired their disease by injection, most of which were administered at pre-
natal visits. Only two of the male cases age 15-29 years acquired disease through this mode of
transmission.

Investigation revealed that parenteral injection was the principal mode of administration of
nearly all medicines at the mission hospital. Each morning five syringes and needles were issued
to the nursing staff for use at the outpatient department, the prenatal clinic, and the inpatient
wards. These syringes and needles were apparently not sterilized between use on different
patients but rinsed in a pan of warm water instead. At the end of the day they were sometimes
boiled. The surgical theater had its own supply of instruments, syringes, and needles, which were
stored separately and autoclaved after use.

Virus transmission was interrupted by stopping injections and by isolation of patients in their
villages. Use of protective clothing and respirators, strict isolation of hospitalized patients,
and careful disposal of potentially contaminated fomites may also have helped in controlling
the epidemic.

The virus responsible for this illness was named Ebola virus, after the name of a small river a
few kilometers from Yambuku. Ongoing studies in Zaire, Sudan, Central African Republic and
Cameroon are investigating the nature of this disease, which since 1976 has been shown to be
endemic in the Congo river basin. An epidemic of Ebola disease occurred in southern Sudan in
1979 and consisted of 33 confirmed cases with a case fatality rate of 67%. The low frequency of
person-to-person transmission of Ebola virus, along with the high mortality rate, indicates that
the agent probably has an animal or some other reservoir in nature, although animals and insects
appeared to have no role in transmission during the epidemic.

15
The International Commission was disbanded on January 1977 following an investigation
involving hundreds of persons and costing more than one million dollars. The following
recommendations were made to the Government of the Republic of Zaire:

1. Maintain active national surveillance for acute hemorrhagic disease. Require regular
positive and negative reporting. Investigate all suspected cases and take appropriate action
including collection of diagnostic specimens, institution of clinical isolation procedures,
and the use of protective clothing for medical personnel.
2. Distribute pertinent information to medical and other personnel participating in
surveillance and update this material by appropriate documents.
3. Organize a national campaign to inform health personnel of the proper methods for
sterilizing syringes and needles in order to ensure that patients are not infected with
diseases from other patients as a result of poor technique.
4. Maintain a list of experienced Zairian personnel so that the appropriate action can be
taken without delay in the event of a new epidemic.
5. Maintain a stock of basic medical supplies and protective clothing for use of
suspected outbreaks.
6. Keep plasma from immune donors in readiness and obtain further information concerning
the effectiveness of this treatment.

References

Baron RC, McCormick JB, Zubeir OA. Ebola virus disease in southern Sudan:
hospital dissemination and intrafamilial spread. BULL WRLD HLTH ORG 1983:
61: 997-1003.

Heymann DL, Weisfeld JS, Webb PA. Johnson KM, Cairns T. Berquist H. Ebola
hemorrhagic fever: Tandala Zaire, 1977-1978. J INFECT DIS 1980; 142: 372-376.

Johnson KM, Scribner CL, McCormick JB. Ecology of Ebola virus: a first clue? J
INFECT DIS 1981; 143: 749-751.

Report of an International Commission. Ebola haemorragic fever in Zaire, 1976. BULL

WRLD HLTH ORG 1978; 56: 271-293.

Saluzzo JF, Gonzales JP, Georges AJ, Johnson KM. Mise en evidence d.anticorps vis a vis du
virus de Marburg parmis les populations humaines du Sud-Est de la Republique Centrafricaine.
C.R.ACAD.SCI. (Paris) 1981; 292: 29-31.

LeDuc JW. Epidemiology of hemorrhagic fever viruses. Review Infect Dis 1989; 11 :S730-735.

Sureau PH. Firsthand clinical observations of hemorrhagic manifestations in Ebola

hemorrhagic fever in Zaire. Reviews Infect Dis 1989:11;S790-S793

16