The survey design was a cross-sectional, cluster sample, health facility survey undertaken in government and non-government public health facilities in 15 states of Sudan. Private health facilities were not included in the survey because they are not supplied with ACTs and malaria diagnostics by the Federal Ministry of Health. The survey was undertaken between 9th and 31st December 2009. The sample size was calculated at patient level assuming 50% prevalence of the primary case-management indicator, 95% confidence level, desired precision of +/- 7%, design effect of 2 and the likelihood that 50% of health facilities would not have either malaria diagnostics or ACTs in stock. Based on these parameters, an estimated sample size was 728 outpatient consultations in each age group (below and above 5 years of age) or estimating that on average 3 febrile patients can be recruited in each age group at each facility over one survey day, then the required number of surveyed facilities was 243 (728/3).
From the universe of 5,716 public health facilities in Sudan a stratified random sample taking into consideration within-country distribution of facilities by state (15 strata) and type of facilities within a state (3 strata) was drawn to ensure national representativeness. For each of 45 strata we calculated a sampling fraction within the universe of 5,716 facilities, proportional and effective sample size within the required sample of 243 facilities, and finally undertook simple random sampling for each stratum. The final effective sample size contained one additional facility because of rounding off to the proportional sample size. A cluster was defined as all outpatients' encounters between a health worker and patients occurring on a single survey day.
Data were collected by 72 field workers trained during 5 days until agreement of practice results between field workers and trainers was greater than 90%. In total, 36 teams were created, each composed of two field workers. The field work was supervised by 15 state coordinators who were also trainers of field workers. Three methods were used to collect data during a survey day at health facility. First, all patients presenting to the outpatient departments underwent rapid screening when they were ready to leave the facility. After obtaining written informed consent non-referred and non-pregnant patients aged 2 months and older, weighing 5 kg and above and presenting for an initial outpatient visit with fever were interviewed. Detailed information about patients' age, weight, temperature, prior use of antimalarial drugs, routine malaria diagnostics performed and results reported, medications prescribed and the key counseling and drug dispensing tasks performed during the facility visit was collected during the interviews and from the patient-held cards. Second, each health facility was assessed to determine availability of basic infrastructure, equipment, job aids, antimalarial drugs, malaria RDTs and functional malaria microscopy service. Finally, all health workers who attended to recruited patients on the survey day were interviewed at the end of the survey day to determine their demographics, pre-service and in-service training, access to treatment protocols and exposure to supervision in the preceding 6 months. Verbal informed consent was obtained from all health workers prior to the commencement of study procedures.
Indicators and definitions
The key study indicators referred to the availability of malaria case-management commodities, health workers coverage with ACT related interventions and the malaria case-management practices which are deemed critical to the success of the AS+SP based case-management policy. The primary indicators at health facility level referred to the proportions of facilities with ACTs, other antimalarials, job aids and malaria diagnostic services on the survey day. At the health worker level the primary indicators were the proportions of health workers who received training on malaria case-management, who had access to national treatment protocols and who were exposed to supervisory visits including malaria case-management in past 6 months.
At the patient level primary indicators referred to febrile patients for whom malaria testing was merited and AS+SP treatment should have been considered in relation to the result of malaria test. With respect to testing indications, we included all patients with fever because national malaria guidelines are ambiguous and do not specify criteria defining other obvious causes of fever (Figure ). Therefore our analysis included febrile, non-pregnant patients aged 2 months and older, weighing 5 kg and above and presenting for an initial outpatient visit without being referred or admitted for hospitalization. Fever was defined as axillary temperature of ≥ 37.5°C or the history of fever during the present illness. A health workers decision to refer or admit patient for hospitalization was used as proxy measure to define severity of disease and exclude patients with suspected severe disease meriting treatment with other drugs than AS+SP. The primary outcome was a composite case-management indicator and included performance of all of the following three tasks: 1) patient was tested for malaria; 2) if positive test result patient was treated with AS+SP, and 3) if negative test result patient was not treated for malaria. Despite some ambiguity of the guidelines with respect to the treatment of patients with test negative results (Figure ), the absence of antimalarial treatment for these patients was included in the definition of the composite case-management indicator because treatment of test negative patients was considered as one of the major determinants undermining the new case-management policy. The secondary outcomes reflected individual components of the case-management including testing, treatment, dispensing and counseling practices in various patients' subgroups.
The general plan for malaria diagnosis and treatment in Sudan.
Data management and analysis
Data were double entered by independent data entry clerks using Access 2007 (Microsoft Inc., Redmond, WA, USA). Data files were compared for errors using a verification programme and referring to original questionnaires. The analysis was performed using STATA, version 11 (StataCorp, College Station, TX, USA). Descriptive analysis was undertaken at the health facility, health worker, and patient level across all age groups and stratified for children below 5 years of age and patients 5 years and older. The presentation of the results reflected the main objective of the study, i.e. to provide national level data for which the study was sufficiently powered to obtain information with desired precision. A stepwise approach was applied in the analysis of malaria case-management practices. First, to evaluate progress of 2007-2012 national malaria strategy health workers practices are analyzed at all health facilities regardless of the availability of the case-management commodities. Second, to evaluate health workers adherence to the new policy the same analysis was restricted to the facilities where AS + SP and malaria diagnostics were available on the survey days. Finally, the quality of AS + SP dispensing and counseling practices was restricted to patients who had AS + SP prescribed and dispensed at facility. The precision of proportions (95% confidence interval [CI]) was determined adjusting for the cluster sampling at health facility level. Cluster adjusted chi-square test was used to compare proportions between patients below and above 5 years of age.
Ethical clearance for the study was obtained from the National Research Ethics Committee of the Federal Ministry of Health in Sudan (reference number 114-11-09).