ACT has been recommended as a life-saving intervention for malaria cases especially in sub-Saharan African countries which claims more than 60% of the global malaria morbidity [30, 31]. However, full benefits cannot materialize until important health system elements, including patients’ access, are addressed. It is argued that malaria elimination, an essential step towards the long-term goal of eradication, can only be achieved when a high proportion of patients with malaria have access to effective treatment . This is critical for the malaria eradication agenda. As is the case for many other endemic sub-Saharan African countries, malaria is generally on the decline in Tanzania [1–8]. However, a new push is required to sustain the gains and move towards malaria elimination. It is in this way that this bottleneck needs urgent attention if the potential of ACT is to be optimized and this control agenda is to be fulfilled. It is unfortunate for malaria control initiatives that when substantial resources are being invested in the development and procurement of efficacious drugs, these tools do not reach the patients in a timely way that would contribute to malaria reduction commensurate with their demonstrated capacity. ACT is most effective against malaria when acting on parasites in cases not yet severely complicated .
This is one of few studies that have demonstrated patients’ access to ACT within the context of system effectiveness framework [25, 26] in two neighbouring HDSS sites located in three districts in Tanzania. Although the results are not representative of the country composed of more than 130 districts, they offer an insight on the health system’s performance and an estimate of timely access to ACT. As they are HDSS sites, their populations are investigated more often and several health system interventions were implemented on a research basis than elsewhere in the country [7, 18, 21, 22, 27, 28]. Hence it is reasonable to estimate that these study results should not be worse than the rest of Tanzania. The country’s latest Tanzania HIV/AIDS and malaria indicator survey (2011-2012) was a nationwide study, the results of which suggest 21% access to malaria treatment within 24h . Observations from this study and those reported from TDHS 2011 fall short of 80% target set by Roll Back Malaria for achieving the potential public health benefits of ACT . They suggest that more than half of people ill with malaria in Tanzania do not have timely access to quality ACT and are exposed to risk of poor outcomes of the disease. These results are comparable to the rest of rural Africa where malaria transmission is endemic where a similar picture of malaria patients’ access to ACT has been reported in Burkina Faso and Kenya [30, 34].
Fever patients’ access to ACT has long been investigated in KU , following its implementation of the Access Programme from 2004 to 2007 , which involved social marketing campaigns that stressed the importance of prompt and effective treatment for malaria. The KU site was also among the first districts in the country having ADDOs. There were 44 ADDOs in the study area at the time of our survey giving an outlet to population ratio of 0.4 per 1,000. Rufiji was characterized by four ADDOs as the study was in progress with an outlets to population ratio of 0.05 per 1,000. Hence, both social marketing and ADDOs should have increased timely access to appropriate treatment in KU compared with Rufiji but we did not detect a significant difference. This could be interpreted that these interventions did not have an impact on fever patients’ timely access to ACT although access after 24 hours was better in KU.
Timely access to ACT previously observed in KU from the ACCESS programme was above 60% . This is better than the rate reported in the present study. This discrepancy would be explained in two ways. One, it would be associated with different dimensions used to report effective coverage between our surveys and study conducted during the ACCESS project. Alba et al. presented actual anti-malarial drug taken by patients within 24 h following fever episode as reported by patients or by their caretakers. Patients would take drugs available at home or from a neighbour or friend and this has been reported in several studies [15, 31, 35, 36]. Wider availability of sulphadoxine-pyrimethamine (SP) and older drugs during ACCESS survey was reported from KU . Thirty percent and 16% of patients self-treated at home from Rufiji and Kilombero-Ulanga respectively from our study (data not shown in the results section). In this study, results presented reflect patient or caretaker reported visit made to ACT providers within 24 h on the onset of fever illness. As reported in the preceding sections, results on which this paper is based have been generated from just one access dimension of effective coverage. Other components proposed in the effectiveness framework (targeting accuracy, prescribers’ compliance and patients’ adherence)  have been conducted using different designs whose results are published separately . Findings from these papers will demonstrate whether the patients obtained and used the medicines comparable to ACCESS publications. On the other hand, the social marketing campaigns and ADDOs roll out might have improved timely treatment access for fever as demonstrated in surveys that monitored their effects. And there would be marked difference in timely access between the intervention and comparator site. However, it is also possible that these effects were not sustained over time.
The study has shown more than 15 percentage points statistically significant difference in 48 h patient access between KU and Rufiji sites. The larger ACT outlets population ratio associated with roll out of ADDOs in KU compared to Rufiji could be singled out as an important reason for this difference. It would imply that after the introduction of ADDOs patients have become closer to treatment providers and their number making treatment outlets visits has increased but many of them have not yet started seeking care within 24 h. People still keep delaying to seek treatment from appropriate providers. Observations from some other studies conducted in one of these study areas has suggested an association in this delay with initial actions taken at home with anti-pyretic and locally available anti-malarial medicines .
A logistic regression model was run to determine whether there was an association in the observed access with any factors usually estimated as risks for low patients’ access to malaria treatment. This form of analysis assumes that risks of exposure to malaria especially in high endemicity areas and chances for treatment-seeking are not normally distributed . The burden is concentrated in the poorest quintile of the population whose composition is dominated by small children, women and those living in remote rural communities. Hence, malaria control interventions can achieve intensive and broader effectiveness if their coverage is equally clustered around these groups highest at malaria transmission in high burden countries . The risks of inherent effectiveness decay on the poor highlighted in the equity effectiveness loop framework would be addressed when this is properly handled . This study has not been able to prove whether this is happening. Women, under-fives and the poorest of the poor were shown to have almost similar access to ACT providers within 24 hours of malaria onset (Table 4). However, the study cannot claim definite lack of difference in timely access to treatment outlets in these communities by age group or gender as it was not powered to test this. Nevertheless, in a study among the very poor in Tanzania based on observations made in the same area, Schellenberg et al. demonstrated unequal access to appropriate treatment between the poorest and the least poor in that area . This difference could have been addressed by subsequent implementation of interventions that used pro-poor approaches [19, 42], such as rolling out of ADDOs, and it might also have happened because many public health interventions were implemented in the area generating duplicative effect that has extended to every population group. As was stated earlier, Kilombero-Ulanga and Rufiji HDSS sites were a testing ground for many public health intervention studies in Africa. However, the only condition that the study was powered to demonstrate was access within 24 hours according to malaria transmission season.
This study’s exploration of patients’ access to ACT providers was based on an assumption that fever could be a robust indicator of malaria. Fever has actually been an entry point for several studies seeking to understand treatment-seeking practices for malaria including patients’ access to treatment providers. This has been the case because the widespread adoption of Integrated Management of Childhood Illnesses (IMCI) that encouraged treatment of every fever episode as malaria at a time when malaria transmission was so endemic in Africa and the likelihood of being infected with malaria parasites, especially for small children, was so high. Challenges posed by diagnostic technology in existence at the time and the need to treat patients as timely as possible to prevent progression of the disease into severe form was an important argument raised to justify IMCI strategy. However, as malaria transmission decreases across sub-Saharan Africa, it has been demonstrated that the presence of rapid diagnostic tests is making IMCI no longer an attractive strategy for malaria. Fevers are now increasingly associated with other diseases than malaria, including bacterial infections and especially pneumonia [38, 43]. Hence, an assumption that all fevers are malaria is becoming increasingly inappropriate.