Pertussis outbreak investigation of Dara Malo district, Gamo Administrative Zone, Southern Nations, Nationalities and Peoples Region, Southern Ethiopia


 Background: The aim of this outbreak investigation was to verify and describe a pertussis outbreak in the Dara Malo district of Ethiopia. Method: a descriptive cross-sectional study was conducted using district Public Health Emergency and Management (PHEM) surveillance data and outbreak management field reports. Stratified attack rates and fatality for pertussis are described. Systemic problems leading to the outbreak are explored and narrated. A modified CDC pertussis case definition was employed with polymerase chain reaction used to confirm cases. Results: From September 2018 to January 2019, 1840 suspected, probable and confirmed pertussis cases and six deaths were identified. Pertussis cases ranged from 1 month to 51 years in age and outbreak occurred in 14 out of the 24 villages of Dara Malo district. The overall attack rate was 1708 per 100,000 population with a fatality rate of 3.3 per 1,000 pertussis cases. The highest attack rate of 12,689/100,000 was seen in infants. Investigations suggest low vaccine coverage and cold chain management system leading to low vaccine potency as the most likely reason for the outbreak. in these community were: low pentavalent three vaccination coverage and non- functional. In addition, poor implementation of surveillance programs, poor health seeking behavior of the community may have led to delayed and suboptimal responses to the outbreak. Conclusion: There is an urgent need to build capacity to strengthen routine vaccination services capacitate health workers to manage outbreaks. In addition, improving the maintenance of cold chain must be prioritised. Other LMICs, are urged to take lessons learnt from this outbreak and strengthen their own vaccination programs.


Background
Pertussis is an extremely infectious vaccine preventable disease caused by the gram-negative coccobacillus, Bordetella pertussis and less commonly by Bordetella parapertussis [1]. Pertussis remains endemic and has reemerged as a public health problem in many countries despite decades of high vaccination coverage [2].
Pertussis is widely distributed in many countries throughout the world. Globally, around 24.1 million pertussis cases and 160,700 deaths from pertussis were reported in children younger than five years in 2014. The African region contributed the largest proportion with 7.8 million (33%) cases and 92,500 (58%) deaths.
Infants and young children have remained most susceptible to pertussis-related morbidity and mortality. Around 5.1 million (21%) estimated pertussis cases and 85,900 (53%) estimated deaths were in infants younger than one year [3,4].
In children, pertussis presents classically with paroxysms of cough ending with the characteristic whoop and post-tussive vomiting. However, in young infants, pertussis can initially present as apneic or cyanotic episodes prior to development of cough [5]. The mode of transmission is person-to-person via aerosolized respiratory droplets or by direct contact with respiratory secretions. In its early catarrhal stage, pertussis is highly contagious, with a secondary attack rate of up to 90% among non-immune household contacts. Untreated patients may transmit infection for up to three weeks or more following the onset of typical coughing attacks, although communicability diminishes rapidly after the catarrhal stage [6].
The most common complication, and the cause of most pertussis-related death is secondary bacterial pneumonia. Young infants are at the highest risk for acquiring pertussis-associated complications [7]. Adolescents and adults are an important reservoir for B. pertussis and are often the source of infection for children [8,9].
Vaccination is the best way to prevent pertussis in all age groups [10]. Evidence suggests that high coverage with highly efficacious vaccines leads to high levels of protection in children under five years of age. In contrast, even minor reductions in overall coverage can lead to an increase in cases [11]. Completion of a three-dose pertussis-containing vaccine schedule prevents 80% of cases and 95% deaths.
Incomplete immunization has been shown to prevent severe morbidity with one dose and two doses decreasing mortality by 50% and 80%, respectively [12,13].
In Ethiopia, a pertussis whole cell vaccine (wP) has been in use since 1980 [14].
From 2007 this has been in the form of a pentavalent that in addition to pertussis whole cell vaccine, also contains Diphtheria, Tetanus, Hepatitis B and Haemophilus influenza type B (DTP-HepB1-Hib1) vaccines. The pentavalent vaccine is given three doses scheduled at 6, 10 and 14 weeks of birth [14,15].
Based on World Health Organization (WHO) estimates the Ethiopian national coverage of one dose of pertussis containing vaccines in 2018 was 85% while that of three doses was 72% which is less than the average coverage for Sub Saharan countries [10]. This immunization coverage differs greatly between districts and regions of Ethiopia with coverages of 3 doses of pentavalent ranging from 20.1% in Afar to 81.4% in Tigray. The difference in vaccine coverage is partly due to geographical inaccessibility and other social determinants [16]. However, according to the routine health management information system (RHMIS) report, Dara Malo district health office reported a higher vaccine coverage for the third pentavalent dose to be 103.0% in 2016/2017 and 94.8% in 2017/2018. The pentavalent vaccine dropout rates were 3.0% and 3.4% respectively over the same period [17].
Uncoordinated and poor epidemic response mechanism is a challenge in containing epidemics, reducing morbidity, mortality and economic loss in different countries.
Diagnosis of pertussis, which is largely based on clinical picture, may be modified by age, history of previous immunization or infection, antibiotic exposure and concurrent infection with other pathogens [18]. According to WHO vaccinepreventable diseases 2019 global summary report there were no reported pertussis cases from 2000 to 2017 in Ethiopia. In 2018, however, there were 2423 cases of pertussis reported in Ethiopia [15]. WHO data is further corroborated by independent data sources that indicated the occurrence of cases of pertussis in Ethiopia [19,20].
This study focused on describing an outbreak of pertussis in the Dara Malo district of Gamo Administrative Zone, Southern Ethiopia, and also investigate possible factors that may have led to the outbreak of pertussis.

Study area and population
Dara Malo is one of 160 districts administered under Southern Nations, Nationalities and Peoples Region (SNNPR), Ethiopia. The district has 24 administrative structures consisting of one urban and 23 rural villages [17]. Based on the 2007 Ethiopian national census, the projected population of Dara Malo district in 2018 was 107,715 with 16,814 under five and 3,436 under one years of age [17].

Study design and sampling methods
The study employed a facility based cross-sectional descriptive study design. Data were sourced from the Public Health Emergency and Management (PHEM) database that includes surveillance data on pertussis. Available data included information on age, sex, geographical location, immunization status, dates of onset and health facility visits, treatment received, and outcomes of reported pertussis cases. In addition, the study reviews availability data to explore the cause of the outbreak in terms of functional cold chain, vaccines and supplies system at the targeted health facilities and district health office.
All surveillance records collected from 1 st September 2018 to 9 th January 2019 were reviewed to identify all documented suspected, probable and confirmed pertussis cases during the period. Health records were reviewed for each of these cases.

Data collection
In the pertussis case review component of the study, socio-demographic variables, immunization related information of the children, dates of onset and health facility visits, treatment received, and its outcome will be addressed from the line list.
The data were collected by four trained health professionals, each with a master's degree in public health (MPH) and had experience of reviewing documents. In addition, project advisors and program officers from Transform: Primary Health Care (TPHC), a USAID program, oversaw the quality of collected data.
The quality of data was assured first by providing training to the data collectors and their supervisors. At the end of every data collection day, a meeting was held between the data collectors and a supervisor to discuss practical problems and issues of major concern. The investigators rechecked the completeness of the collected data on a daily bases.

Case definition
To confirm pertussis nasopharyngeal samples were taken from suspected cases by experienced laboratory professionals from the Ethiopian Public Health Institutes and tested using polymerase chain reaction (PCR) for Bordetella pertussis and Bordetella parapertussis (RealStar® Bordetella PCR Kit 1.0. altona Diagnostics). Suspected cases with paroxysmal cough illness for more than 2 weeks were assessed for possible epidemiological link to cases confirmed by a positive PCR for Bordetella pertussis. Pertussis cases reported to district health office PHEM department were classified by the outbreak investigation team as laboratory confirmed, probable or suspected pertussis using modified Center for Diseases Prevention and Control (CDC) criteria [21].

Confirmed:
A confirmed pertussis case is defined as acute cough illness of any duration with a positive PCR for B. pertussis, or a case that meets the clinical case definition and is epidemiologically linked directly to a lab-confirmed case [22]. In this study, confirmed pertussis case is an acute cough illness of any duration with a positive PCR for B. pertussis.

Probable:
A probable case is defined as cough lasting ≥ 2 weeks AND paroxysms of coughing, inspiratory "whoop" or post-tussive vomiting AND No laboratory confirmation AND No epidemiologic linkage to a lab-confirmed case. In this study, probable pertussis case is a case Cough lasting ≥ 2 weeks AND Paroxysms of coughing, inspiratory "whoop" or post-tussive vomiting AND No laboratory confirmation [22].

Suspected:
A suspected case is defined as a non-improving cough of 14 days or more or cough of any duration with paroxysms or cough of any duration with whoop.

Epidemic of pertussis:
According to CDC pertussis epidemic is defined as a situation when two or more cases clustered in time.
≥2 PCR confirmed cases clustered in time (within 42 days of each other) and space (e.g. in one building) where transmission is suspected to have occurred in that setting (e.g. nosocomial transmission in a hospital) OR 9 following: paroxysms of cough, inspiratory whoop, or post-tussive vomiting [22].

Data management and analysis
The data were collected in a spreadsheet, cleaned, summarized and analyzed using Microsoft Excel® and Statistical Packaged for Social Sciences Research (SPSS IBM V 20) [22].
Proportions of cases have been presented as percentages. All continuous data were summarized using medians and ranges. The frequency of cases was reported as a case attack per 100,000 population using the 2018 population for Dara Malo as the denominator, while deaths were reported as fatality rates per 1000 cases. Epidemic curves were depicted using date of onset of paroxysmal cough and date cases were identified by health workers. This analysis shows the propagated nature of the outbreak. Data were reported stratified by age groups and geographical location, and described in terms of time, place and person.
The result of the quantitative analysis is presented in frequency tables and graphs.
In addition, the findings of filed reports and observation were thematically analyzed to explain possible factors for the reported pertussis outbreak.
In addition, using ArcView GIS (Ver 10) a spatial analysis was made to show most affected villages using attack rates and case fatality rates.

Ethical clearance
Ethical clearance was obtained from both JSI Research & Training Institute, Inc. and the SNNPR Health Bureau, Ethiopia, Institute Review Boards (IRBs). Permission to conduct the study was obtained from facility managers. The PHEM core process established and deployed epidemic investigation team which comprised a field epidemiologist, an integrated diseases surveillance response officer and health extension workers. The team members were oriented on the case definitions, diagnosis and treatment protocols, strengthening community-based surveillance, and enhancing community mobilization strategies.

Verifying the pertussis outbreak
The index case , a five-year-old child with unknown vaccination status was identified Over the whole surveillance period a total of 1840 suspected, probable and confirmed pertussis cases including six deaths, were identified.

Descriptive analysis of Reported Pertussis Outbreak
The pertussis outbreak line-list is described in terms of time, place and person.  Table I. Dara Dime village reported 838 (45.5%) of the pertussis cases. The age of the cases ranged from 1 month to 50 years of age with a median of 36 months. Of the total cases, 230 (12.5%) were aged less than six months while 801 (42.5%) were 1 to 4 years old Table 1.

Demographic distribution of the cases
The overall attack rate was 1,708/100,000. The attack rates by age category were 12,689/100,000, 5,965/100,000 and 667/100,000 reported in under one-year infants, 1 to 4 years children and 5 or more years of children and adults, respectively (additional file 1).
A total of six deaths due to pertussis were reported to give a total case fatality rate of 3.3 deaths per 1000 cases. These include five deaths from Daro Dime and one death from Menena Aba kebele. Two deaths each were under 6 months, and between 6 to 12 months. One death each were from 1 to 4 years and 5 to 9 years.
The highest case fatality rate was found among infants less than 5 months of age with a case fatality rate of 8.7/1000 cases, followed by infants aged 6 to 11 months with 4.85/1000 cases ( Figure 1).

Clinical presentation
The most frequent presentation was paroxysmal cough which was a complaint in all 1840 (100.0%) cases. This was followed by post-tussive vomiting which was

Geographical distribution of the outbreak cases
The highest attack rate was reported from Dara Dime village while the highest case fatality rate was noted in Menena Aba village. Figure 4 depicts ArcView GIS spatial analysis pertussis case attack rates and case fatality rates by village.

Vaccination status
The vaccination status of pertussis cases was collected from reviewed EPI registers at health posts and oral report of family members or care takers. Number of cases who had received a pertussis-containing vaccine was found to be 169 (9.2%), 321 (17.4%), and 761 (41.4%) for one dose, two doses and three doses, respectively. Of the rest of reported pertussis cases 232 (12.6%) had not received any vaccine dose while 357 (19.4%) had unknown immunization status.

Management of the pertussis outbreak
There were delays in notification and lack of timely proper management of probable or suspected pertussis cases in the community. The health workers were treating cases as pneumonia, pertussis was never suspected until the death of four children occurred due to the outbreak.
The outbreak response team lobbied community leaders and elders to mobilize the community to report cough of 2 weeks or more to nearby health posts. Information leaflets were developed and distributed to all household members using local language, Gamo.
Individual cases were treated with antibiotics to reduce severity and duration of symptoms, and to prevent complications. All schools were closed for the period of mass preventive prophylaxis campaign while community members were advised to avoid participating in local markets. In addition, community members were advised to implement strict personal hygiene including frequent hand washing and avoiding contact with pertussis suspected individual.
Following the last reported case, the District Health Office continued active case search for more than two months on daily bases. The outreach investigation team continued to provide support for individual case management, mass health education for community members and enhance the capacity of health workers on cold chain management, defaulter tracing, and use of data for decision making.
After no further cases of pertussis were reported from 10 th January to 27 th March 2019, the investigation team was demobilized.

Environmental factors
The housing condition and family size per household was assessed. The average house in affected community had 7 people and lacked air circulation. Community members were encouraged to open ventilate their homes.

Cold chain, immunization services management
Wacha Health Center has high tension electric power supply. Three out of four health center have solar panel to get un-interrupted electric power for cold chain maintenance. All fourteen health posts were collecting vaccine on scheduled vaccination campaign and outreach services. The investigation team did not find continuously recorded temperature monitoring tools. And the observed vaccine vial monitor (VVM) which is a thermochromic label on vials, revealed that the vaccines were kept at temperatures which do not preserve vaccine potency. Overall cold chain maintenance was generally found to be poor.

Discussion
This study confirms a pertussis that occurred between September 2018 and March 2019, in Dara Malo, a district in Southern Ethiopia. Pertussis outbreak cases occurred in 14 of 24 villages in the district and involved 1840 cases with six deaths.
Pertussis is one of the under-reported major causes of morbidity and mortality among children in many low-income countries [3]. This report confirms the importance of strengthening community-based active surveillance programs and the need to enhance the capacity of community health workers in identifying pertussis suspected cases as well as providing prompt effective response to any outbreaks [23,24]. The current outbreak could have been the result of poor vaccination coverage combined with lack of standardized national pertussis responsiveness and management guidelines.
Though the district health office confer the presence of high rate of community based surveillance weekly report on public health notifiable diseases for over one year, the health systems did not identify the unusual health condition occurred in Dara Malo district. This finding might be occurred due to limited capacity of health care system in identifying risks, developing emergency preparedness plan, diagnosis and treatment of individual cases, prompt response and management capacity.
The signs and symptoms for both uncomplicated and severe cases were consistent with either of CDC and WHO pertussis case definitions. proposed to report for pertussis suspected cases and the cause of this outbreak was Bordetella pertussis, which was laboratory confirmed causative agent using PCR. The overall pertussis case attack rate of 1,708 per 100,000 inhabitants of Dara Malo district was much higher than previously documented pertussis attack rate of 130 per 100,000 population reported in South Wollo, North Eastern Ethiopia in 2017 [19] or the 400 per 100,000 population reported in Papua New Guinea in 2012 [25]. The high attack rate documented in this investigation could be due to the delay in appropriate management control with nationally recommended and effective antibiotic treatment for individual cases [26]; contact tracing was largely incomplete and the intervention did not include prophylaxis antibiotics administration to susceptible community members until late in the outbreak. Failure to protect close contacts of both confirmed and probable pertussis cases might have encouraged the spread of infection over an extended period. The low health seeking behaviour of the community (with more than two months of delays recorded to visit the health facility) most likely due to lack of knowledge on severity and risk of pertussis in the community would have served to exacerbate the situation. As expected, the highest attack rate was in infants followed by children aged one to four years. This finding was consistent with the report of Yeung et al's global estimates [3] and the report of the outbreak in the South Wollo area of Ethiopia [19].
The overall pertussis case fatality rate of 3.3 per 1000 pertussis cases was much lower than the 37 per 1,000 pertussis cases from Mekdela [19] and the 30 per 1,000 pertussis cases reported from Papua New Guinea (2012) [25]. Four of the six deaths were reported before outbreak control interventions were initiated. The early initiation of the responsive program may have averted further deaths. All cases identified during the outbreak were treated with antibiotics which consisted of amoxicillin, and cotrimoxazole, erythromycin or azithromycin. The use of antibiotics would have reduced the possibility of severe pneumonia which can occur as a complication of pertussis infection [24].
The cause of this outbreak could be multifaceted. Even though, the RHMIS reported a high rate of vaccination coverage of over 95% with three doses of a pertussis containing vaccine in 2016/17 for Dara Malo district., a large proportion of pertussis cases were not fully vaccinated. The inconsistence between routine RHMIS outbreak data and may be due to poor data recording. Maintaining high vaccine coverage is one of the most important interventions in reducing pertussis cases and consequently averting outbreaks [11]. The poor vaccine vial monitoring observed during the outbreak investigation and non-functional cold chain management system at primary health care units may have resulted in poor vaccine potency even in cases that received all doses [29,30] Limitations The study was limited by lack of resources to confirm the majority of suspected cases, however, in an outbreak, clinical criteria have been shown to have high sensitivity and specificity in the diagnosis of pertussis [31]. In addition, this study did not capture the reason for delayed health seeking behaviour of care takers and the completeness, consistency and reliability of immunization data. As immunization records were not always available, information on vaccination status could have been affected by recall bias.

Conclusions
There is an urgent need to build capacity to strengthen routine vaccination services including defaulter tracing to keep vaccine coverage at high level to avert outbreaks. Vaccinating all children with appropriate doses for age must be supplemented with building and strengthening the capacity of health workers to timeously and properly manage outbreaks when these occur. In addition, static and outreach services, community based surveillance activities and maintaining the cold chain need to be prioritised. A review of data quality status and capacity to respond to outbreaks at district level must be urgently undertaken.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.   Figure 1 Pertussis cases fatality rate per 1000 cases by age group at Dara Malo district, Gamo Zone, Figure 2 Bar chart presenting reported sings of severity among pertussis cases, cough in Dara Malo D Figure 3 Epi Curve of Pertussis cases by date of onset of paroxysmal cough and identified by health w