A case-control study to assess risk factors related to cholera outbreak in Addis Ababa, Ethiopia, July 2016

Introduction Suspected cholera cases were reported to the city administration health bureau in Addis-Ababa, Ethiopia on June 5, 2016 and it was laboratory confirmed by culture on 7 June 2016. The outbreak was declared by the government on June 8, 2016. The aim of this study was to rapidly investigate the outbreak epidemiologically and guide response activities in the affected Sub cities of Addis Ababa city administration. Methods An unmatched 1:2 case-control study was conducted in six selected sub-cities of the city administration. Different laboratory tests were also done from suspected possible risk factors identified to support the epidemiological findings. A case was a patient greater than 5 years old, who developed acute watery diarrhea with or without Vomiting. Control was an individual greater than 5 years' old who stayed in the same township and did not suffer from cholera. A structured questionnaire was used to select cases and controls. Epi InfoTM statistical software was used to analyze the data. Results were presented in narratives, figures and tables. Results The present study found that, the study participants who used street-vended water (Odds Ratio (OR)=10.4; 95% CI: 1.20-90.9), those who reported holy water sources use (OR=2.7, 95% CI: 1.45-5.04), eating raw meat (OR=7.75; 95% CI: 3.86-15.54) or roasted meat (OR=2.16; 95% CI: 1.19-3.93) and vegetable salad (OR=2.07; 95% CI: 1.14-3.76) were associated with contracting cholera infection. The likelihood of contracting cholera was significantly higher among those who ate food from street vendor sources (OR=5.32; 95% CI: 1.82-15.56) and those who practiced open defecation (OR=8.12; 95% CI (2.20-29.81). Having a latrine (OR=0.29; 95% CI: 0.12-0.69) and proper hand hygiene practice (OR=0.22; 95% CI: 0.14-0.38) were found to be protective against cholera infection. Conclusion Improving awareness of the community by intensifying social mobilization activities through community participation in proper hygienic practice, proper waste disposal and latrine facility construction and utilization. Provision of safe water for the community by strictly conducting end-point assessment of water points and conducting a KAP assessment among food handlers to explore their knowledge and practices regarding safe food/drink handling and water treatment as well as initiate appropriate PH actions based on the findings which will be necessary for prevention of similar cholera outbreaks in the future.


Introduction
Acute watery diarrhea is the passage of three or more loose or watery profuse diarrhea per day, in a period of 2 hours to 5 days. Other types of diarrhea include dysentery; and persistent diarrhea which lasts 14 days or longer. Patients presenting with diarrhea can also present with other symptoms such as vomiting, fever, and body weakness.
Diarrhea causes loss of body fluid and electrolytes, which can result in dehydration. If dehydration is not corrected, death can be the result [1]. The causes of diarrhea can be bacterial, viral or parasitic.
Bacteria causes of diarrhea include vibrio cholera, escherichia coli, campylobacter jejuni, salmonellae and shigella species. Viral causes of diarrhea include rotavirus, adenovirus, and coronaviruses. Parasitic causes of diarrhea include giardia, entamoeba, cryptosporidium and the helminths (Strongyloides, Schistosoma) [2]. Among these causative agents, acute watery diarrhea caused by vibrio cholerae serogroup O1 and O139 is the highly pathological [3,4]. It is usually transmitted through faecally contaminated water or food and remains an ever-present risk in many countries [5]. Vibrio cholerae produces a toxin that stimulates the secretion of water and electrolytes in the intestinal tract. Patients with cholera may suffer from acute watery diarrhea, vomiting, and dehydration but rarely present with fever [6].
In Ethiopia it was indicated that, there was a cholera epidemic in 1990 which persisted with recrudescence of cases till 1998 [7]. Moreover, from July 2008 to June 2009 in Ethiopia, there were a total of 9485 cases and 193 deaths (with case-fatality rate 2.0%) of acute watery diarrhea in six regions including Addis Ababa [8]. Addis Ababa city is one of the city administrations of Ethiopia, according to the Central Statistical Authority report, the total population of the city administration is 3,384,569 [9]. During 2001, a total of 4000 cases and 40 deaths of cholera with a total case fatality rate of 0.33% were reported from the city administration.

Statement of the problem and rationale
The first index case was detected and reported from private clinic Rapid assessment was conducted by the team deployed to the city administration with the aim of understanding the extent of the outbreak and identifies possible risk factors for the outbreak to support the preparedness towards implementing prevention and control activities. The assessment was done by visiting treatment sites and interviewing cases in CTCs and interviewing health professionals working at the CTCs and respective health system structures by using questionnaire and checklist prepared by the team. Possible risk factors implicated from initial assessments include drinking water from unprotected sources; rivers, springs and holy water sites; Open defecation due to lack of latrines, poor solid waste collection and disposal, Poor food hygiene, and overcrowding. Different factors including, the onset of the rainy season could lead to further spread of the outbreak. Besides this, large-scale population movements in and out of Addis Ababa, the capital city of Ethiopia increases the potential of further spread to other parts of the country and outside the country. Rapid identification of risk factors which were contributing for the outbreak was with paramount important to support the prevention and control activities for the outbreak.
General objective: to determine the factor(s) associated with increased risk of infection with cholera and to direct/guide/refine ongoing cholera outbreak prevention and control activities with a view to stopping transmission.  Suspected case: a case of cholera should be suspected in: [10] an area where cholera is not known to be present, a patient aged 5years or more develops severe dehydration or dies from acute watery diarrhea; in an area where there is a cholera epidemic, a patient aged 5 years or more develops acute watery diarrhea, with or without vomiting. At the health post and at community levels, a suspected cholera case can be defined as follows: [10] any person 5 years of age or more with profuse acute watery diarrhea and vomiting Confirmed case: a suspected case in which vibrio cholera O1 or O139 has been isolated from their stool [10].
Sample size determination (assumptions): least extreme odds ratio to be detected =2, hypothetical proportion of exposed controls =40%, hypothetical proportion of exposed cases =60%, confidence level =95%, Power =80%, number of cases =150, number of controls =150, total sample size =300 (Kelsey et al.) [11]. Selection of cases: the primary study subjects were newly confirmed cases of cholera admitted to the Cholera Treatment Centre (CTC) in the selected 6 hot spot sub-cities during data collection time.
Line list in CTCs were used to identify newly admitted cases and systematically select cases for interview. Systematic selections of cases (every two cases from the line list) were used to ensure the random allocation of cases during the selection process. This principle was applied until the required numbers of cases were obtained. Cases were selected from the CTCs in six hotspot sub-cities during the study Exclusion criteria: to maintain specificity, children under-5 was not included in this study. Persons who had recently recovered from cholera were also excluded.
Data collection: following the selection of cases and controls, and receipt of informed consent, cases (or their caregiver where the case is too ill to speak) and control subjects were interviewed using the questionnaire in appendix A and data were collected on some identifying information, demographic information, and date of onset and exposure to risk factors for infection including water and food sources, within three days prior to onset of disease. Two field epidemiology residents for each sub-city totally 12 residents were recruited as a data collector and 1 WHO surveillance officer was allocated for each sub-city totally 6 officers as supervisors for all hot spot sub-cities. Both the data collectors and supervisors were given 1-day orientation on their duties and responsibilities during the data collection process. Data processing and analysis: using un-matched case-control design, the study examined the differences in types of exposures between individuals with confirmed and epi-linked cholera and healthy controls in order to determine the risk associated with that exposure. A standard questionnaire was used to collect information.
The purpose of this case-control study was to determine if a case was more likely to have been exposed to a risk factor in the three days before onset of illness than a similar uninfected person in the community. The study was done prospectively, enrolling new cases in the Cholera Treatment Centers (CTC) as they were identified. The controls who were persons not sick with cholera were chosen at random from the same community of residence of the cases. Data entry and analysis was conducted using the Epi-Info version 7.2. The associations between risk factors and infection among cases and controls were analyzed using chi-square statistics and expressed as odds ratios with 95% confidence intervals.

Epidemiological Findings
Demographic characteristics of cases and controls: a total of 300 (100 cases and 200 controls) study participants participated in this study making a response rate of 100%. The age and occupation of the cases and controls were comparable but gender was not (corrected for during analysis). Most of the case respondents 60 (61.2%) and control respondents 143 (72.6%) were between 15-44 years age group. Almost 22% (22) of the cases and 20% (38) of the control's respondents were between 45 and 64 years of age. The median age for study participants of cases was 37 (range 11-80) and controls were 30 (range 10-90) ( Table 1). The occupational status of the case respondents was also comparable with controls ( Figure 2). Clinical presentation of cases: most of the cases wich were brought to the CTC center for treatment were brought with acute watery diarrhoea (99%), vomiting (80%), dehydration (67%) and others sign like fever and joint pain (4%) ( Table 2).

Clinical presentation of cholera cases, July 2016, Addis Ababa
Drinking water exposure: cases and controls were also compared with the drinking water exposure. Drinking street vended water and holy water were significantly associated with developing cholera.
Those individuals who drank street vended water were 10.47 (1.20-90.90) times more likely to develop the infection than who did not with a p-value of 0.017. Similarly, those individuals who drank from holy water were 2.7 with CI of (1.45-5.04) more likely to develop the infection than those who did not drink with a p-value of 0.001 (Table   3). On the contrary treating water before drinking has a protective effect. Those who treated water before drinking had 0.17 (0.09-0.28) times protected than those who did not treat water before drinking with a p-value of 0.000 (Table 3).

Food and beverage exposures of cases and controls: in regard
to food and beverage exposures of cases and controls, eating raw meat, partially cooked meat and eating vegetable salad were significantly associated with developing the disease. Cases were more likely than controls to eat raw meat, partially cooked meat and vegetable salad. Those individuals who ate raw meat, partially cooked meat and vegetable salad had 7.75 (CI-3.86-15.54), 2.16 (CI-1.19-3.93), 2.07(CI-1.14-3.76) times more chance of developing cholera than their counterparts with a p-value of 0.000, 0.016 and 0.015 (Table 4). (1.36-5.32) times more likely to develop cholera than those who did not eat with a p-value of 0.001 and 0.003 (Table 5).
Hygiene and sanitation: in regards to Hygiene and sanitation practices among cases and controls, cases were more likely to practice open defecation, having a latrine and hand hygiene were a protective effect. Those individuals who practice open defecation were 8.12 (CI-2.20-29.81) times more likely to develop cholera than those who did not practice open defecation. Those who used the latrine, washed hands before eating and defecation were 0.29 (CI-0.12-0.69), 0.32(CI-0.19-0.52) and 0.22 (CI-0.14-0.38) more protective effect than those who did not with a p-value of 0.004, 0.000 and 0.000 (Table 6).

Discussion
As of 4 th August, the cumulative number of cases reported was 5.879 (AR=0.17%) and 12 deaths (CFR 0.20 %). Among them, 64% were males and 36% were females. All the 10 sub-cities and all Woredas in the sub-cities had reported cases. Most of the cases affected by this outbreak were from the 15-45 age groups, greater than 60% of the cases admitted and treated were within this age group. 139 indicates that there was also person to person transmission of the disease from diseased patients through above listed risk factors. This type of epidemic pattern for this outbreak was recorded in many previously affected countries like Haiti and Zimbabwe [13,15]. In this study, drinking street vended water and Holy water was a statistically significant risk factor of contracting cholera. This finding is comparable with the findings of cholera outbreak investigation in Sierra Leone and watery diarrhea in Zimbabwe by the year 2012 and 2011, respectively [12,13]. Both cases and controls had very high access to improved water sources, with public taps representing the most common type of improved water source, but the residual chlorine in several spot checks of public taps were found below the minimum requirement level. This finding suggests that either chlorine introduced into the water system at its origin was not reaching some water points in the area sampled, or that chlorination may not have been occurring.
On the other hand, sewage has entered networks of water pipes through cross-contamination with sewer pipes in some areas. The lack of chlorine and unhygienic water storage at household level could make this and other improved water sources susceptible to contamination with disease-causing agent like vibrio cholerae.
Overcrowding, competing priorities and limited resources have resulted in poor hygiene and sanitation conditions. These conditions combined with low chlorination rates create an environment highly vulnerable to cholera transmission. An evaluation of the integrity of water distribution networks and chlorination practices would be useful to identify and resolve potential deficiencies. In the mid-long term, improvement and expansion of the water and sanitation infrastructure will help to prevent future epidemics of cholera and other waterborne diseases. Regarding sanitation practices, this study revealed that those individuals who practice open defecation were 8.12 times more likely to develop cholera than those who did not practice open defecation. Besides this, using latrine regularly, washing hands before eating and defecation had shown more protective effect. This study finding on sanitary practice was compatible with other studies done in different countries. Several studies have shown the health benefits of hand washing with soap and water and using latrine has significant role in the prevention and control/spread of communicable diseases [16][17][18]. In this study, a wide range of food items were tested to see if they were associated with the risk of cholera infection.
The collected data indicate that eating raw meat products (like raw meat and dullet) and vegetables was a risk factor for cholera infection. Even though this study was not designed to determine whether street vendors had the knowledge or supplies to practice sanitary food and drink preparation, this area deserves further exploration. Information, Education and Communication (IEC) campaign focused on the importance of safe food handling practices and the importance of consuming water that has been treated with a chlorine product are core public health prevention activities.

Conclusion
The age and occupation of the cases and controls were comparable but the gender was not (corrected for during analysis). Statistically significant risk factors for cholera found in this study include: drinking street vended water and holy water, eating raw meat, partially roasted meat, vegetable salad or unboiled fresh milk; food consumption at a street vendor or restaurant, lack of access to latrine (open defecation), most common type of water source is tap water, but end-point water quality needs to be closely monitored (eg residual chlorine testing) and evidence that population is responding to household water treatment and hand washing advice. We will recommend that they: conduct repeated end-point water sampling and testing including holy water sites to ensure improved water quality, reinforce point-of-use drinking water treatment to ensure safe drinking water, conduct a KAP study among food handlers to explore their knowledge and practices regarding safe food/drink handling and water treatment, and initiate appropriate PH actions and intensify social mobilization campaigns on hand washing, water treatment and sanitation.
What is known about this topic  Causes for Acute Watery Diarrhea can be bacterial, viral or parasitic;  Regular monitoring at a point in use for drinking water and food and drink establishments should be given higher priority.

Competing interests
The authors declare no competing interests.

Acknowledgements
Our enormous gratitude goes to Addis Ababa Health bureau staffs, St Paul Millennium Medical College field epidemiology residents, EPHI staffs and WHO experts for their contribution during the data collection and data management. We want also to thank Addis Ababa Health Bureau and Ethiopian Public Health Institute for their willingness to conduct the study and use the findings for the prevention and outbreak control. Lastly, we want to acknowledge the investigation team and study participants for their unreserved support and participation during the study.