No abstract

Expected value of information methods evaluate the potential health benefits that can be obtained from conducting new research to reduce uncertainty in the parameters of a cost-effectiveness analysis model, hence reducing decision uncertainty. Expected value of partial perfect information (EVPPI) provides an upper limit to the health gains that can be obtained from conducting a new study on a subset of parameters in the cost-effectiveness analysis and can therefore be used as a sensitivity analysis to identify parameters that most contribute to decision uncertainty and to help guide decisions around which types of study are of most value to prioritize for funding. A common general approach is to use nested Monte Carlo simulation to obtain an estimate of EVPPI. This approach is computationally intensive, can lead to significant sampling bias if an inadequate number of inner samples are obtained, and incorrect results can be obtained if correlations between parameters are not dealt with appropriately. In this article, we set out a range of methods for estimating EVPPI that avoid the need for nested simulation: reparameterization of the net benefit function, Taylor series approximations, and restricted cubic spline estimation of conditional expectations. For each method, we set out the generalized functional form that net benefit must take for the method to be valid. By specifying this functional form, our methods are able to focus on components of the model in which approximation is required, avoiding the complexities involved in developing statistical approximations for the model as a whole. Our methods also allow for any correlations that might exist between model parameters. We illustrate the methods using an example of fluid resuscitation in African children with severe malaria.

BackgroundChildren presenting to hospital with recent or current Plasmodium falciparum malaria are at increased the risk of invasive bacterial disease, largely enteric gram-negative organisms (ENGO), which is associated with increased mortality and recurrent morbidity. Although incompletely understood, the most likely source of EGNO is the bowel. We hypothesised that as a result of impaired gut-barrier function endotoxin (lipopolysaccharide), present in the cell-wall of EGNO and in substantial quantities in the gut, is translocated into the bloodstream, and contributes to the pathophysiology of children with severe malaria.MethodsWe conducted a prospective study in 257 children presenting with malaria to two hospitals in Kenya and Uganda. We analysed the clinical presentation, endotoxin and cytokine concentration.ResultsEndotoxaemia (endotoxin activity ≥0.4 EAA Units) was observed in 71 (27.6%) children but its presence was independent of both disease severity and outcome. Endotoxaemia was more frequent in children with severe anaemia but not specifically associated with other complications of malaria. Endotoxaemia was associated with a depressed inflammatory and anti-inflammatory cytokine response. Plasma endotoxin levels in severe malaria negatively correlated with IL6, IL10 and TGFβ (Spearman rho: TNFα: r=−0.122, p=0.121; IL6: r=−0.330, p<0.0001; IL10: r=−0.461, p<0.0001; TGFβ: r=−0.173, p<0.027).ConclusionsEndotoxaemia is common in malaria and results in temporary immune paralysis, similar to that observed in patients with sepsis and experimentally-induced endotoxaemia. Intense sequestration of P. falciparum-infected erythrocytes within the endothelial bed of the gut has been observed in pathological studies and may lead to gut-barrier dysfuction. The association of endotoxaemia with the anaemia phenotype implies that it may contribute to the dyserythropoesis accompanying malaria through inflammation. Both of these factors feasibly underpin the susceptibility to EGNO co-infection. Further research is required to investigate this initial finding, with a view to future treatment trials targeting mechanism and appropriate antimicrobial treatment.

BackgroundHypoglycaemia is an independent risk factor for death in severe malaria and a recognized adverse treatment effect of parenteral quinine. In 2006 our hospital changed quinine treatment policy from 15 mg/kg loading (plus 10 mg/kg 12-hourly) to 20 mg/kg loading (plus 10 mg/kg 8-hourly) to comply with new WHO guidelines. This presented us with the opportunity to examine whether there was any dose relationship of quinine and hypoglycaemia occurrence.MethodsRetrospective case notes review of all children admitted to hospital with severe falciparum malaria between April 2002 - July 2009, before and after the introduction of the new WHO quinine regimen. Four-hourly bedside glucose levels were measured until intravenous quinine was discontinued. Clinical events immediately preceding or concurrent with each episode of hypoglycaemia (glucose < = 3.0 mmol/l) were recorded.Results954 children received the old quinine regime and 283 received the new regime. We found no evidence of an increased prevalence of hypoglycaemia (< = 3.0 mmol/L) on the new regime compared to former (15% vs. 15%); similar findings were noted for profound hypoglycaemia (< 2.2 mmols/L) 8% v 5%, P = 0.07. Episodes were co-incident with disease severity markers: coma (57%), circulatory failure (38%) and respiratory distress (21%) but less commonly with seizures (10%). Disruption of maintenance fluids and/or blood transfusion concurred with 42% of the hypoglycaemia episodes. Post admission hypoglycaemia increased odds of fatal outcome (24%) compared to euglycaemic counterparts (8%), odds ratio = 3.45 (95% confidence interval = 2.30-5.16) P < 0.01.ConclusionThere was no evidence to indicate a dose relationship between quinine and occurrence of hypoglycaemia. Hypoglycaemia concurred with severity features, disruption of glucose infusion and transfusion. Careful glucose monitoring should be targeted to these complications where resources are limited.

Efficiency gains is a potential strategy to expand Kenyas fiscal space for health. We explored health sector stakeholders understanding of efficiency and their perceptions of the factors that influence the efficiency of county health systems in Kenya. We collected data during a stakeholder engagement workshop. Workshop participants included health sector stakeholders from the national ministry of health and 10 (out 47) county health departments, and non-state actors in Kenya. We divided stakeholders into three groups and carried out facilitated group discussions followed by whole group feedback and discussion session. A total of 25 health sector stakeholders participated. We analysed data using a thematic approach. Health sector stakeholders indicated the need for the outputs and outcomes of a health system to be aligned to community health needs. They felt that both hardware aspects of the system (such as the financial resources, infrastructure, human resources for health) and software aspects of the system (such as health sector policies, public finance management systems, actor relationships) should be considered as inputs in the analysis of county health system efficiency. They also felt that while traditional indicators of health system performance such as intervention coverage or outcomes for infectious diseases, and reproductive, maternal, neonatal and child health (RMNCH) are still relevant, emerging epidemiological trends characterized by an increase in the burden of non-communicable diseases (NCDs) should also be considered. The stakeholders identified public finance management, human resources for health, political interests, corruption, management capacity, and poor coordination as factors that influence the efficiency of county health systems. An in-depth examination of the factors that influence the efficiency of county health systems could illuminate potential policy levers for generating efficiency gains. Mixed methods approaches could facilitate the study of both hardware and software factors that are considered inputs, outputs or factors that influence health system efficiency.