The study is a double blind randomised controlled clinical trial with two parallel groups. The trials is designed to be compliant with the CONSORT Statement 
All patients who present with a history of yellow oleander poisoning will be assessed to determine if they are eligible for the study. Those who meet the criteria will be approached to give their written informed consent, following which they will be randomised (Figure 1). Patients who do not initially meet the criteria will be reviewed regularly and approached if they meet the inclusion criteria at a later time point.
patients who are admitted to adult medical wards, who are over the age of 16years, and who have any of the following manifestations of oleander induced cardiac toxicity:
Patients with documented ischemic or valvular heart disease, pregnant women and age under 16 years.
Standard oleander patient management guidelines will be followed. These are based on the national poisoning treatment guidelines. The only difference between enrolled patients and those not enrolled will be the addition of FDP/placebo intervention.
Trial intervention and study procedures
Patients randomised to the treatment arm will be treated with 250mg/kg loading dose of FDP (Esafosfina from Biomedical Foscama, Italy) over 20minutes followed by 6mg/kg/hr for 24 hours in addition to standard care. Patients randomised into the control arm will be treated with an equal volume (equal to the volume of FDP in the treatment arm) of 0.9% saline as a bolus and a 24 hour infusion. All attending doctors and nurses will be blinded to the treatment.
Clinical parameters such as systolic and diastolic blood pressure will be monitored for 24 hours. A Holter monitor will record the cardiac rhythm for 48 hours. All cardiac events will be recorded. If a serious cardiac rhythm abnormality recurs after an initial response to the bolus (within 2 hours of a bolus), a further bolus of 250mg/Kg of FDP (or equal volume of placebo) may be given at the discretion of the treating physician while the infusion will be maintained at the same rate.
Randomisation is done using purpose designed computer software. The random sequence and allocation are concealed prior to randomisation. The program will randomise eligible patients in a 1:1 ratio. The allocation sequences are generated and encrypted independently by an IT consultant who has no role in patient recruitment, treatment and assessment.
The randomization will be performed by study pharmacists centrally. If a patient meets the inclusion criteria and gives consent, clinical research assistants will call the pharmacist with details such as name, hospital number and weight of the patients. Then the pharmacist will randomise the patients and prepare the placebo or active treatments for the study team. The allocation will only be known by the pharmacists who will have no other role in patient management and data collection. Intention-to-Treat analysis will be applied. That is, the analysis will include all randomised patients in the groups to which they were assigned, regardless of non-compliance, protocol deviations, withdrawal, and anything that happens thereafter 
The primary outcome of this study is the reversion to sustained sinus rhythm with a heart rate >50 bpm within 2 hours of completion of bolus.
Secondary outcomes include:
Change from baseline serum potassium on the 6, 12, 18 and 24 hour blood samples.
Time to revert and duration of reversion to sinus rhythm on the Holter monitor over the first 24 hours (reflecting the efficacy of bolus and infusion)
We are planning to recruit 240 symptomatic patients with yellow oleander poisoning admitted to Teaching Hospital, Kurunegala and National Hospital Sri Lanka, Colombo. In the study by Eddleston M et al , 2/32 (6.25%, 95% CI 1% - 19%) patients reverted to sinus rhythm spontaneously after 2 hours. Second degree and third degree heart block are markers of toxicity but may not be strong predictors of death (which may be due to myocardial toxicity rather than atrio-ventricular conduction block). Reversion to sinus rhythm is more likely with less severe second degree block. We would ideally have a sample size that ensured a significant number of third degree blocks which are less likely to revert spontaneously so that the effect of treatment in this sub-group can be seen. However, the proportions of the different grades of block are not well quantified and thus precise estimates of power in sub-groups are not possible.
We expect up to 15% spontaneous reversion by two hours and believe a 40% reversion in the active treatment arm would be clinically significant and likely to provide strong evidence that this would translate to a mortality benefit. This would require just 60 patients in each arm (α = 0.05, β = 0.8, missing data/dropout 10%). To increase the likelihood of recruiting enough patients with 3rd degree heart block to see if the effect is observed in the most severely poisoned patients, and to account for possible differences between centres we intend to do a study of 120 patients in each arm.
Study hypothesis and principal comparisons
The primary outcome of the study is to investigate if the addition of FDP (250mg/kg loading dose over 20 minutes followed by 6mg/kg/hr for 24 hours) to routine treatment will completely reverse serious arrhythmias within two hours. The study will also investigate the effect of FDP on the presence of abnormal rhythms and serum potassium and other electrolytes over 24 hours.
The Primary outcome (proportion with reversion to sinus rhythm) will be compared with the chi-squared test. Kaplan Meier curves will be constructed to demonstrate the cumulative reversion to sinus rhythm over time. A Cox proportional hazard model will be used to test the overall difference in reversion to sinus rhythm between treatments adjusted for baseline variables as necessary. A longitudinal statistical technique known as Generalised Estimating Equation (GEE) will be used to analyze changes in electrolytes (e.g. potassium) over time.
Independent data monitoring and ethics committee (IDMEC)
The IDMEC will conduct the interim analysis and examine primary and secondary outcomes.
We intend to do a planned interim analysis once a total of 120 patients had been randomised (i.e. 60 in the treatment arm and 60 in placebo) which may lead to modification of or cessation of the trial as outlined below. The IDMEC will be asked specifically to comment on the need for subsequent modification of the trial protocol for the infusion. An infusion rate modification should be recommended if there is strong evidence of a response to the bolus which is not sustained by the infusion. An interim analysis at this stage will also be used to stop the trial if there is very strong evidence for efficacy. We have a 95% power to detect a 60% treatment effect over the expected placebo response of 15% without any loss of power for the overall study (α = 0.001, β = 0.95, missing data/dropouts 10%). If such a major response is noted the IDMEC will instruct the trial team to stop the trial.
This study is approved by the ethics research committee of University of Peradeniya and the Human Research Ethics Committee of the Australian National University. Written informed consent will be obtained from all patients in their native language (Sinhala or Tamil).