INTRODUCTION: Systemic sepsis occurs in 20-50 % of patients admitted to non-cardiac intensive care units and despite management, continues to be a serious clinical challenge. Development of novel therapeutic strategies is hampered by lack of an animal model which accurately and reproducibly mimics the clinical condition. AIMS: To refine a rat model of caecal ligation and puncture (CLP) induced sepsis; to investigate the effect of CLP on the structure of selected regions of the central and peripheral nervous system; to investigate the effect of CLP on structure and function of the gastrointestinal tract. METHODS: Under anaesthesia, the caecum was either ligated below the ileocaecal valve and punctured or only mobilised and the animals were allowed to recover for 18- 20 hours, during which time the following parameters were investigated: bacteraemia; locomotor activity; food and water intake; body weight; core temperature; serum lactate and glucose; white blood cell and platelet cell counts. The spinal cord, sciatic nerves and gut tissues were taken for quantitative and qualitative analysis using light microscopy. The above parameters were also investigated in a group of un-operated rats. In vitro studies were performed to compare the spontaneous and evoked contractile activity of the stomach, duodenum, colon and hepatic portal vein from either CLP or sham CLP animals. RESULTS: CLP but not sham CLP induced bacteraemia, lethargy (depressed nocturnal locomotor activity), anorexia, gastric retention, weight loss, pyrexia (within 4-5 hours) followed by hypothermia (from 17 hours), leucopaenia, thrombocytopaenia and hypoglycaemia after 18-20 hours. CLP also induced injury to the spinal cord and sciatic nerve characterised by peri-vascular oedema and myelin sheath swelling. There was massive mucosal sloughing, mucosal haemorrhage and vascular congestion in the gut tissue, indicative of gastrointestinal dysfunction in CLP but not in sham CLP rats. No significant differences were found between un-operated and sham CLP animals. Although CLP inflicted severe damage to the gut mucosa and inflammatory cell infiltration of the muscle layers, gastric, duodenal, colonic and hepatic portal vein tissue exhibited spontaneous contractile activity, although generally at a frequency lower than in sham CLP animals and with an elevated baseline tone. In addition, tissues were still capable of responding to a range of pharmacological agonists including acetylcholine, 5- hydroxytryptamine, phenylephrine and thrombin although the responses were modified in CLP animals. CONCLUSION: The spectrum of functional and histopathological changes described in the current study following CLP in the rat are consistent with the development of multi-organ dysfunction syndrome observed in patients with sepsis and argue that the refined model described is more clinically relevant for investigation of novel therapies than other models (e.g. LPS infusion). The results allow a reduction in the number of animals used by obviating the need for an un-operated group and permit the reduction of the post-CLP observation time required to obtain valid pathological results compared to previous investigations. The pathological changes observed in the spinal cord and sciatic nerve following CLP could explain the acute myelopathy and critical illness polyneuropathy that frequently occur following sepsis. The in vitro gut studies provide an insight into the mechanism(s) underlying gut dysfunction observed in patients with sepsis and show that despite severe damage the gut is likely to retain its capacity to respond to prokinetic drugs which could be used to treat motor disorders. These findings raise the possibility of identifying therapeutic interventions to restore neural and gastrointestinal tract function to normal and promote patients recovery

Most in-hospital deaths secondary to infections in under-five deaths within sub-Saharan Africa (SSA) occur in the initial 24 hours of admission and shock has been identified as a major risk factor for the early deaths. However, controversies exist on the appropriate clinical diagnosis of shock, choice of ideal fluid for resuscitation (crystalloid or colloid), and safety of fluid resuscitation in severe malnutrition or severe malaria. This thesis investigates these aspects and also reviews the evidence base of current paediatric fluid resuscitation guidelines for children (aged >60 days and ≤12 years) with severe febrile illnesses. Capillary refill time >2 seconds, weak pulse volume, or bradycardia, in the presence of abnormal temperature and severe disease are predictive of impaired perfusion (defined by lactic acidosis) and death. Tachycardia and temperature gradient are neither associated with increased risk of death nor predictive of hypoperfusion. Existing international definitions of shock have low sensitivities (FEAST=44%, WHO=2%, and ACCM=59%) and high specificities (FEAST=82%, WHO=100%, and ACCM=66%) for diagnosis of impaired perfusion. Clinical criteria derived (called derived shock) had a sensitivity of 30% and specificity of 93%. Shock in children with severe febrile illnesses in Kilifi has a complex presentation but mainly presents with hyperdynamic circulation (high cardiac index) and vasodilatation. Cases with low cardiac index (myocardial dysfunction) are relatively rare but increase the risk of mortality when present. Synthetic colloids (gelofusine, hydroxyethyl starch 130/0.4 (HES), and dextran 70) are safe for use in fluid resuscitation in children with severe malaria. However, HES is the most promising compared to other synthetic colloids concerns still remain about its renal safety. However, further evaluation of synthetic colloids for treatment of shock is not warranted due to the findings of FEAST trial. A Pilot trial shows that bolus isotonic fluids are safe, have better efficacy, and produce faster resolution of shock compared to low-sodium solutions at volumes and rates recommended by WHO in children with severe malnutrition. Evidence available from all ten the trials in children with sepsis show that fluid resuscitation using crystalloids and colloids result in similar survival. However, fluid bolus resuscitation results in increased mortality compared to no bolus (control) in children in SSA. This finding excludes children with gastroenteritis, trauma, burns, and malnutrition. Colloids are better than crystalloids for severe dengue shock but both have similar efficacy in moderate dengue shock.