Malaria, as a public health priority, is an important focus point for development of quality pharmaceutical services, and can serve as a model for other endemic diseases in tropical areas of the world. This study encompassed a special period for the PNCM, since the 2001 guideline was being reviewed and new ACT schemes were being introduced for P. falciparum.
The assumptions in studying pharmaceutical services in high-risk municipalities were simple. First, the Brazilian control strategy is based on diagnosis and treatment, making pharmaceutical services essential. Second, pharmaceutical services must be organized in those endemic diseases areas where health needs are critical. Third, it is important to study highly endemic areas, as services are apt to be more structured and organized that in others where prevalence is low and the disease is not perceived as a health risk. Finally, the choice of municipalities was intentionally made to include not only incidence, but different population magnitude, to reflect municipal diversity in the Brazilian Amazon.
Although choice of municipalities may be perceived as a limitation of this study and may have implications on generalization of results, especially for rural areas, the results may be applicable to urban areas in the Brazilian Amazon. Furthermore, the effort was made to give a comprehensive view of organization of services and of prescribing in municipalities of various population sizes once it was estimated that population size was a good predictor for services organization .
Results showed that P. vivax malaria is the most prevalent species found in the visited municipalities, repeating the pattern observed in the Brazilian Amazon . Manaus and Ariquemes, which had been high-risk municipalities in 2005, were excluded from this category in 2007 (Table 1). Number of cases dropped considerably in all municipalities in 2008. However, only these two maintained their high-risk-free status.
It was observed that 80% (12/15) of facilities had an available copy of the 2001 official guideline for malaria treatment and this is a positive indicator of quality of care. Nevertheless the guideline itself had problems, since it was based on limited pharmaceutical and clinical evidence . In 2010, the PNCM published new guidelines with better quality and based on more consistent and locally-generated scientific evidence .
Additionally, 87% of health services had other instructional material. However this information had not been disseminated to all the health workers in the area . An especially critical issue in rational medicines use is availability of information, for users and for health professionals alike. Uninformed health professionals may be unable to treat patients adequately. Reports show that with appropriate information even patients with little education were able to follow treatment with artemether-lumefantrine, without supervision [19, 20].
Another positive aspect was that no expired anti-malarials were found. Nonetheless, overstock was detected and none of the municipalities adhered to forecasting methods of stock management. None of the health facilities fulfilled 100% of good practices in storage and dispensing areas. As was to be expected, re-supply of stocks followed no discernible routines, in spite of a perceived sense of urgency in making treatment available. Forecasting activities are an essential part of pharmaceutical services organization. This aspect is tantamount, especially in the Amazon, where humidity and temperatures are very high most part of the year. In fact, a sampling of anti-malarials from Brazilian treatment facilities was submitted to quantitative analytical assays and presented problems in concentration of the active principles . The evidence suggests that if in urban areas this situation was detected, in rural regions, where facilities may be less than adequate, with deficiencies in storage and dispensing areas, where staff is possibly lacking and large distances may cause supply shortages, the situation should be worse, as has been cited in the literature [22, 23].
In respect to diagnostic coverage, this study showed that all individuals were submitted to laboratory confirmation of malaria and that in only 24 patients (4%) species was not identified and that these were given empirical treatment. In spite of reflecting only urban areas, this operational indicator showed good performance, following WHO recommendations (at least 90% of suspected malaria outpatient cases should undergo laboratory diagnosis-thick smear or Rapid Diagnostic Test) . Different results are reported from some African countries, where laboratory diagnostic coverage revolved from 33% to 40% [25, 26], while in other settings 60% of patients have no access to diagnosis .
Maybe the failures in diagnosis were due to probable test error. Albeit considered simple and less technologically intense that other laboratory methods, the thick smear may be a tricky technique. It is prone to considerable inter-observer variation, while requiring continuous training of staff, as well as minimum laboratory infrastructure . Brazil has recently introduced a new career for health workers, that of "microscopy technician", to give professional incentives to health technicians working in malaria , decreasing rapid turnover of staff in endemic areas and consequently favoring diagnostic quality.
The findings of this study show that patients have close to immediate access to microscopic diagnosis. The average time between diagnostic and treatment was 0 days. This was considered an excellent result, considering other examples in the literature . In the state of Amazonas 57% of patients are offered treatment within 48 hours of symptoms. In Rondonia and in Acre, the proportions are 65% and 71% respectively. The national Brazilian average is 59% . The findings may reflect the urban scenario depicted in the study.
Complete regimens for both P. vivax and P. falciparum were readily available in the visited facilities. For P. vivax patients, treatment indication according to the national guidelines was successful, while for P. falciparum patients were offered more alternatives. During the study period only one first-line regimen for uncomplicated malaria in adults was recommended for P. vivax. For P. falciparum, the guideline offered two alternatives, because of Plasmodium resistance and a third as consequence of introduction of ACTs in Amazonas.
Nonetheless, a percentage of treatments did not comply with the national guidelines. While one ACT fixed-dose combination was being introduced in Amazonas, another was undergoing clinical field study in Acre - an intervention involving 17, 000 patients and sponsored by The Brazilian Ministry of Health and by Ravreda (Rede Amazônica de Vigilância de Resistência aos Anti-maláricos) . The tentative incorporation of ACT to the treatment guideline was, therefore, not uniform throughout the Amazon. The low number of empirical treatments was due to diagnostic failure and not to inobservance of the national protocol.
Only 5.7% of individuals with malaria received prescriptions or written information on treatment regimens. Only one medical doctor was identified among 63 health workers involved in malaria treatment in the six study sites. These facts reflect part of the malaria care-giving urban scenario in Brazil. Nonetheless, in reference centers for malaria in the Brazilian Amazon, all individuals with malaria are attended to by medical doctors . Brazilian Health System and PNCM policy recommends that better clinical evaluation of each case should be guaranteed through physician-attended care. Therefore, it is difficult to accept that in the visited municipalities a lesser standard of care should be acceptable. Decentralized health interventions have been considered as advances of the Brazilian Health System, in regard to access for users and accountability for municipal managers, but differences in revenue, demographics, health indicator performance and in quality of management result in better or worse malaria control .
In spite of the fact that one of the inclusion criteria was the availability of a prescription or a written instruction to the patient, this was not observed in the field. The existence of a medical prescription is a very important concern since malaria should be treated with a combination of drugs, and the therapeutic scheme depends on the parasitic species, patient age group and clinical severity of the disease. Moreover, many malaria patients have little formal education  and the complexity of the therapeutic regimens for both P. vivax and P. falciparum may baffle them. They rely on family or friends to help with treatment schemes, making a clear readable prescription therefore important to enhance adequate adherence to a complete treatment [34, 35]. The lack of a formal prescription may have, therefore, made the actual treatment regimen given to each patient difficult to detail, since the study relied mostly upon observation and chart reviews.
It is possible to hypothesize that the rest of non-ACT/non-protocol treatments for P. falciparum may have been caused by the lack of medical doctors and pharmacists and by deficient training of other health professionals. Trained staff could nullify or bar inadequate medicines indications. Adherence to therapeutic guidelines goes hand in hand with capacity building of health staff and available information on treatment [22, 25].
Only two thirds of the visited facilities presented a specific area for receiving and consulting patients. In pharmaceutical services, a proper environment for patient dispensing and counseling is a determinant in the patient's understanding of the treatment regimen and of possible adverse effects caused by medicines. Layout of patient facilities is recognized as fundamental in guaranteeing better patient comprehension . This, in turn, may lead to good or bad adherence to treatment .
It is worthwhile to point out that a specific programme for malaria with a separate structure of pharmaceutical services, although helping to focus and prioritize disease needs, may be more difficult to manage. In Brazil, municipalities are responsible for primary healthcare. Although legislation has been passed stressing the importance of combining efforts between primary health care (PHC) and malaria , joint work is absent or lacking in most municipalities. The fact that the PNCM is a vertical programme within the Ministry of Health and centrally managed in regard to procurement is not an excuse for the problems observed in pharmaceutical services. Other vertical programmes have shown a better capacity for integration with PHC . Pharmaceutical services would profit by viewing malaria from the perspective of PHC since all phases, from forecasting to rational use of medicines, apply to both. The literature has observed that difficulties in malaria control are greater where integrations with PHC is lacking .