Final Report on Reported Outbreak of Diarrhoea in Oecussi District, East Timor

The WHO office in Dili first learned of reported diarrhoea related deaths on Wed 19th September. This report was of an increase in cases of diarrhoea and two deaths in infants. Given the background of endemic diarrhoea, both watery and bloody, no immediate action was recommended other than increased surveillance and raising awareness of health service staff.

The following day was a government holiday with no-one available in the MoH. On Fri 21st September, the Ministry was notified of the report; by this time the number of reported deaths had increased to four. Throughout the day staff in the Ministry attempted to get further information from the District and convened a meeting at 1700 on Friday. Shortly before the meeting, WHO received a report from the iNGO operating in Oecussi that the death toll had risen to 21, all in young children.

At the meeting in the Ministry, it was decided that a fact finding visit to the District was essential. The following information was required:

  • Exact size of the reported outbreak in terms of numbers
  • The geographical extent of the outbreak
  • Clarification on the number of deaths
  • Likely cause of the outbreak
  • Adequacy of the District response.

Based on available information, it was postulated that the likely cause was shigellosis, so a supply of nalidixic acid and ciprofloxacin was obtained from the Central Pharmacy to take to the District.

A team from the MoH and WHO visited the District on Sat 22nd September. A report on the visit is at Annex A.

Following this fact finding visit a meeting was convened in the Ministry of Health on Sunday 23 September. This was attended by:

  • The Minister and Deputy Minister of Health
  • The Director General of the Ministry of Health and selected staff
  • Head of Office, Epidemiologist and Laboratory Advisor from WHO
  • Head of Mission of IMC

A report was presented to the Minister and it was agreed that the information available suggested that there was indeed an outbreak of diarrhoeal disease; 400 cases of diarrhoea had been reported in the period between 9 – 22 September. It was noted that although the presumptive cause was shigellosis, clinical experience was that most cases had responded to either cotrimoxazole or chloramphenicol, drugs initially chosen for their availability. It was decided that a team from the MoH assisted by WHO should go to Oecussi to support the District Health Management Team in further investigating and responding to the situation. In view of the limited capacity of the local laboratory services it was further agreed that the assistance of HQ PKF should be sought to collect and test biological specimens. This was subsequently arranged through WHO.

A team comprising one person from MoH, the WHO epidemiologist in Dili and two officers from HQ PKF travelled to Oecussi on Monday 24th, accompanied by the Deputy Minister of Health. The Deputy Minister returned the same day while the team remained for three days.

As previously noted, the District team had responded creditably to the outbreak. All health staff had been alerted and informed of the appropriate management. In light of ongoing clinical experience, initial treatment of cases presenting with bloody continued to be with oral fluids and cotrimoxazole. Other cases of diarrhoea were being appropriately managed with ORS only. The nalidixic acid and ciprofloxacin supplied were being kept in reserve.

Following the previous visit, a form had been prepared to facilitate listing of cases at clinics. This was being distributed by District Health staff. Other agencies had been alerted to the situation and were assisting in different ways as appropriate. Attempts to validate the occurrence and cause of reported deaths were continuing.

It was recommended that an inter-sectoral District Outbreak Response Team be convened in order to manage this situation and to facilitate future emergency/epidemic preparedness activities. This was not supported locally as it was felt it would add unnecessary bureaucracy to the control efforts.

In consultation with the District staff a format was developed for a daily situation report. It was agreed that this be sent to the DHO by 1600 daily in time for a collated report to be sent to the MoH by 1700. A variety of methods were established to facilitate this process including use of CIVPOL and UNMO radio nets, UN DFOs and other international and local staff. Fuller recommendations for case finding and reporting are attached at Annex B.

Having identified that the outbreak was affecting all areas of the district and that deaths had been reported from all sub-districts, a plan was prepared to visit several sites to collect faecal specimens from patients and where possible water samples. The following areas were visited over a two-day period:

  • Makelab clinic and Oenoenu – two house visits were made to patients
  • Baqui clinic
  • Nitibe clinic
  • Mahata and Bobo villages, where a home visit was done and a mobile clinic point visited
  • Pune
  • Bobometo

Faecal and water samples were taken. Faecal specimens were collected mainly by rectal swabbing and immediately transferred to a refrigerated non-nutritive stool transport medium for enteric pathogens (modified Cary-Blair medium, Meridian Diagnostics, Inc. Cincinatti,Ohio). The samples were transferred to the microbiology laboratory at the UN Military Hospital in Dili in insulated, cooled boxes within 24 hours of sampling. Faecal samples were divided between the Military laboratory and the Dili Central laboratory. All water samples were tested initially at the PKF Environmental Health Section and further analysed at the UN Military Hospital microbiology laboratory.

Results

Of eight faecal specimens tested in the UN Military Hospital laboratory, one grew Shigella sonnei; the other seven grew no pathogenic bacteria. A further five specimens sent to the Central laboratory in Dili grew no pathogens. Unfortunately no serotyping facilities are available in East Timor to further identify the Sh sonnei. Antibiotic sensitivity testing revealed that the organism was sensitive to a wide range of antibiotics, much more so than might have been expected. A copy of the report is attached at Annex C.

Eight water samples were tested. All but one were examined for total coliforms and colonies on all plates were too numerous to count. All eight were examined for faecal coliforms; five were contaminated and three had colonies too numerous to count.

Follow-on Report from IMC

A follow up report from IMC, the iNGO operating in the District indicates that after extensive case finding and review of clinic registers, it is estimated that a total of 584 cases of dysentery sought medical assistance during the period 9 – 29 Sept. A significant drop in new cases occurred in the week 30 – 6 Oct. The same report further recorded that 24 deaths due to dysentery had been confirmed! This is in contradiction to previous reports that ascribed only 13 deaths to dysentery. If cases with only diarrhoea and vomiting are excluded, 17 deaths can be ascribed to dysentery.

Comment

Data from the weekly surveillance system operated by WHO, indicate that there has been a total of 3327 cases of diarrhoea reported in Oecussi since the system started collecting data in September 1999; however only 67% of weekly reports were received during this period. The weekly average number of cases reported is 47.5 (95% CI 0-145) and the range from 0 - 341. Increased rates have been observed in the months of June, November and February. Against this background the 400 cases reported in the two weeks from 9 – 22 Sept is high but not outside previous experience.

Over the same period only 9 deaths were reported through this system in all age groups. However, according to IMC Baseline Health Survey Report conducted in Oecussi in Feb-May 2001, <5 mortality is estimated at 3.3 (No CI) deaths per 1000 children per month. This would result in 26.9 expected deaths throughout the district in any given month. The reported 24 deaths in a two week period is approximately double this estimate.

The possibility that this represents excess deaths is increased when it is noted that all 24 deaths were attributed by carers to diarrhoea. However, two possible alternative reasons might explain this; the interviewers were interested in diarrhoea and did not ask about other symptoms and/or due to the publicity surrounding the outbreak, even the slightest diarrhoea was reported in children who died, to the exclusion of other more relevant symptoms.

It is important to clarify this point by establishing the number and causes of deaths during this period in order to establish the true impact of the dysentery outbreak. It is therefore suggested that a verbal autopsy study be carried out in Oecussi District, using the reported deaths as a starting point. This exercise would provide a baseline for Oecussi district and could also serve a as a pilot site for similar surveys in other districts

Conclusion

The conclusion is that this represents a moderate increase in the rate of diarrhoea, which is endemic in this area. The cause is generally poor water and sanitation systems coupled with inadequate personal hygiene practices. It is suggested that the trigger was a deterioration in water quality at the end of the dry season following on a short wet season.

The number of deaths does appear to be in excess of that expected, but given that no confidence limits are available for the estimate of expected deaths, this may also be due to normal variation. Furthermore the exact number and cause of deaths remains unclear.

The isolation of a single shigella indicates that the organism is present and may be responsible for additional cases. The sensitivity pattern indicates that the choice of antibiotic was appropriate and bears out the clinical observations of good response to first line treatments.

Recommendations

It is recommended that:

  • A National Outbreak Response Team be formed in the Ministry of Health with input from the Ministry for Water and Public Works, other relevant Ministries and WHO to prepare plans for managing future outbreaks. A similar Outbreak Response Team should be formed in each District.
  • A survey of the water and sanitation situation in Oecussi be carried out.
  • A Verbal Autopsy survey be carried out in the District to clarify the number and causes of death during the month of September.
  • A national level workshop be convened to discuss the outbreak, the reposnse to it and prepare short and medium term plans to coordinate and improve water and sanitation activities and relevant health education activities in all Districts, but with particular attention to Oecussi.

Annex A. Report on Field Trip to Oecussi District to Investigate reported outbreak of Diarrhoea and Associated Deaths

Following reports on Friday 21 September from Oecussi District of a number of fatalities associated with a bloody diarrhoeal illness in young children, WHO was asked to assist the Ministry of Health in the investigation and control of the situation.

A team from the Ministry, comprising Dr J Martinez Deputy Minister, Dr A Martins Head of Communicable Disease Surveillance and Mr S Da Cruz Head of the Central Laboratory in Dili travelled to Oecussi on Saturday 22 September and was accompanied by myself, Dr SF Drysdale, WHO Epidemiologist in Dili.

A visit was made to the hospital in Oecussi, and several patients were seen and examined accompanied by Dr C Yashimoto. Meetings were held with the District Health Team, the District Administrator and Ms K Crinkleton, IMC Medical Coordinator. A visit was made to an area affected by the outbreak in which there had been several deaths.

IMC made available a report on the outbreak to date, a copy of which is attached. In summary;

  • Real increase in reported cases or diarrhoea first noted on or about 10 Sept.
  • Since then there have been cases reported every day with up to 150 cases reported in a 24-hour period. Using the subsequently agreed case definition, this number reduces by about 50%.
  • There have been 22 reported deaths in young children over the same period; about 13 of these are strongly associated with diarrhoea.
  • Both cases and deaths have been reported form every sub-district in Oecussi.
  • The disease has been affecting almost exclusively (90%) children under 2 years.
  • Features include watery diarrhoea at first, subsequently with pus, mucous or blood; fever; vomiting; variable degrees of dehydration.
  • Working on a presumptive diagnosis of Shigellosis, patients were treated with rehydration, cotrimoxazole. Those who failed to respond were given chloramphenicol, usually to good effect.
  • The District Health Team, greatly assisted by the IMC staff have done sterling work in managing the increased case load and in carrying out health education throughout the District in an attempt to prevent further cases and to ensure those affected are brought to medical attention quickly.

During the meetings it was established that we needed:

  • Further information to describe the extent of the outbreak
  • A working case definition to reduce over-reporting
  • A laboratory diagnosis
  • Recommendations on further control measures

A case definition was agreed: Diarrhoea of 3 or more loose stools in a 24 hour period accompanied by one or more of fever, vomiting or blood/mucous/pus in the stool.

It was agreed to try and collect a line listing of cases by Suko, gathering the following data:

Age

Sex

Village

Date of onset

Presence of vomiting, fever, blood/pus etc

Treatment

Outcome

Drinking water source

Assistance would be sought from PKF, UNTAET CIVPOL and NGOs to collect completed forms.

Two specimens were taken for laboratory testing. Unfortunately one was spilt and may not be usable.

The only additional suggestion on control measures was to distribute empty plastic drinking water bottles widely and to advise people to use solar radiation over a 24-hour period to disinfect their drinking water. This was seen as preferable to starting to distribute chlorine tablets due to the logistic difficulties involved and the fact hat people do not like drinking chlorinated water.

Plans for further visits to the District were left open pending the gathering of further information and the development of the outbreak.

Annex B. Recommendations for reporting cases of dysentery in Oecusse District

To monitor the progress of the epidemic daily collection of data is essential. The data suggested below will allow the district health team to:

  • Determine whether the outbreak is getting worse or better at each clinic
  • Determine how effective the first line treatment is – low proportion of reviews
  • Determine how effective clinic staff are at treating cases – low proportion of referrals
  • Determine how effective the referral system is – no of clinic referrals reaching hospital
  • Severity of the disease – no of deaths; no of hospital admissions

Daily reports should be completed by mobile and fixed clinics, health posts and whatever suko system is set up using the format below. This can be sent by radio using CIVPLO/UNMO radio nets, or by hand, or by phone.

Date ______Clinic ______

Ser / Number of cases / Description
1A / New cases <5
1B / New cases >5
2A / Review Cases <5
2B / Review cases >5
3A / Referrals <5
3B / Referrals >5
4A / Deaths <5
4B / Deaths >5

Similar data should be collected at the hospital daily using the format below.

Date ______

Ser / Number of cases / Description
1A / New cases <5*
1B / New cases >5*
2A / Review Cases <5
2B / Review cases >5
3A / Referrals <5
3B / Referrals >5
4A / Deaths <5
4B / Deaths >5
5A / Admissions <5
5B / Admissions >5
6A / Total number inpatients

* Do not include referred cases; this will result in double counting

The above data must be returned to the District Office by 1500 daily. Here the data will be collated sent to Ministry of Health by 1700 daily. This can be done by phone or by e-mail.

At the District Office an epidemic curve should be constructed for each clinic and for the district as a whole. This will allow the progress of the outbreak to be monitored and appropriate action taken.

On a weekly basis, the line listing form, shown at Appendix 1, should be collected and examined. Daily totals should be adjusted as necessary and the MoH informed of changes.

Any death reported as being due to diarrhoea should be investigated and verified. The fact of death should be confirmed and the cause determined according to questionnaire at Appendix 2.

Appendix 1:Information to be collected on every case of diarrhoea

Case Definition: Anyone suffering from diarrhoea, with 3 or more loose stools in 24 hours with one or more of fever, vomiting or blood/pus/mucus in the stools.

Ser / Name / Village / Age / Sex / Date of Onset / Vomiting / Fever / Blood / Pus/Mucus / Treatment / Hospital?
Y/N / Outcome
Death Confirmed / Yes / No

Appendix 2. Questionnaire to determine cause of death

Date of Interview:______Interviewer:______

Address of Household:______

______

______

______

Name of child:______

Sex:MaleFemale

Name of Main Respondant:______

Relationship to child:______

Child’s date of Birth:______/______/______

Date of death:______/______/______

Where did child die?Hospital

Other health facility

On way to health facility

Home

Other (specify)______

For how long was the child ill before he/she died?______days

Was the child brought to a health facility?YesNoDon’t know

Did the child have a fever?YesNoDon’t know

How many days did it last?_____days

Did the child have frequent loose stools? YesNoDon’t know

(>3 in 24 hours)(diarrhoea)

For how many days?______days

Was there blood in the stools?YesNoDon’t know

Was there pus in the stools?YesNoDon’t know

Was there mucus in the stools?YesNoDon’t know

Did the child have a cough?YesNoDon’t know

Did the child have difficulty breathing? YesNoDon’t know

For how many days?______days

Did the child have fast breathing? YesNoDon’t know

For how many days?_____days

Was the child’s breathing noisy? YesNoDon’t know

Did the child have any fits/convulsions? YesNoDon’t know