Enteric Pathogens and Potential Risk Factors for Acute Bloody Diarrhoea in Nairobi West and Kilifi Sub Counties. A Case Control Study

Abstract

Globally, three million people die due to diarrhoeal diseases every year. Shigellosis is a major cause of diarrhoea-related morbidity and mortality, especially in developing countries, with an estimated annual incidence of 165 million cases and 1 million deaths. Kenya experienced a significant increase in acute bloody diarrhoea cases in Coast, Western, Nyanza and Nairobi regions in 2009 (48,272 cases) and 2010 (64,107 cases). Therefore, it was necessary to determine the epidemiological, clinical and laboratory characteristics of acute bloody diarrhoea cases occurring in the urban and rural populations in Kenya. The study enrolled 805 participants (284 cases and 521 controls) into a hospital based matched case control study between the period of January and December 2012. The main presenting clinical features for bloody diarrhoea cases were: blood in stool (100%) abdominal pain (69%), mucous in stool (61%), loose stools (54%) and anorexia (50%). Pathogen isolation rate from stool was 40.5% (115) with bacterial and protozoal pathogens accounting for 28.2% and 12.3%, respectively. The isolation rate among the rural population (Kilifi) was 24.7% while among the urban population (Nairobi) it was 65.5%. Shigella was the most prevalent bacterial pathogen isolated in 22.8% of the cases while Entamoeba hystolytica was the most prevalent protozoal pathogen isolated in 10.2% of the cases. A total of 86.3% of the bacterial pathogens were resistant to sulfamethoxazoletrimethoprim, 73.8% to tetracycline, 63.8% to ampicillin, 21.3% to chloramphenicol, 3.8% to nalidixic acid, 2.5% to ciprofloxacin and none was resistant to ceftriaxone. High levels of multidrug resistance to three or more antimicrobial agents were observed 68.8% of all bacterial pathogens with resistance in Shigella being 53.7%. There was a serious disconnect between clinical guidelines and clinical practice, clinicians in Nairobi West and Kilifi prescribed to patients 67.6%, 47.7% respectively antimicrobial drugs that were within the high resistance zone (>20%). On binary logistic regression, two factors in rural and three in urban setting remained independently and significantly associated with acute bloody diarrhoea at 5% significance level. In rural setting the factors were: condition of toilet clean and poor general compound cleanliness while in urban setting; other diarrhoea cases in household in previous 2 weeks, drinking water stored in a super drum and hand washing after last defecation were associated with acute bloody diarrhoea transmission. Binary logistic regression for merged rural and urban showed; storage of drinking water separate from water for other use, washing hands after last defecation and presence of coliform in main source water remained significantly associated with acute bloody diarrhoea at 5% significance level. Detection of coliform bacteria in drinking water was used as markers of faecal contamination. A total of 302 water samples {rural (171); urban (131)} were collected and analysed from both settings. In the rural setting, 40.9% of the household water contained total coliforms and 21.5 % faecal coliforms whereas 38.6% of the main source contained total coliforms and 11.9% faecal coliforms. In the urban setting, 10.7% of the household water tested contained total coliforms and 6.2 % faecal coliforms whereas 8.2% of the main source contained total coliforms and 6.2% faecal coliforms. There was a positive correlation between bloody diarrhoea and long term (1971-2013) mean rainfall both in rural (Pearson’s r=0.55) and urban (Pearson’s r=0.85) populations. There was also a positive correlation between bloody diarrhoea and long term (1974-2013) mean minimum and maximum Temp but the correlation with minimum Temp was stronger in rural (Pearson’s r=0.42) and urban (Pearson’s r=0.76).