Apart from measuring the injection prevalence at household level, the study also intended to investigate injection use at health facilities. Indicator If, aimed at establishing injection rates, could be calculated in different ways, depending upon the method which had been chosen to investigate the problem in health facilities, for example, the percentage of prescriptions at a certain health facility that list at least one injection, expressed as:

To establish the percentage of patients receiving an injection when visiting a medical facility, a prescription analysis at the various established medical facilities was performed in Uganda²¹. Prescription analysis for the non-formal providers is not included here since many of these specialize in injections and their injection rates approach 100%. The researchers recorded the diagnosis and treatment of the first 30 patients that visited the provider. This method reveals a high prevalence of injections at all formal health facilities for both regions (Table 3). In 14 formal health facilities visited in Busoga the average prevalence of injection prescription was 68% - this implies that for every ten treatments given, almost seven included an injection. In Ankole, the overall injection prescription rate in 12 provider facilities was 60% - showing that, in every ten treatments given approximately six included an injection. Although injection prescription prevalence for Ankole is lower than that of Busoga (**p=0.001), the overall rates for both regions depict a high prevalence of injection prescription.

²¹A list of the number of the various types of health providers that were included in the Ugandan study is given in Table 12, Appendix 2.B.

In Indonesia, Indicator If may be calculated in two different ways, using either data from the household survey, or from the exit interviews of patients at health facilities (Lebak n=383, Lombok n=391). Indicator If measures here the percentage of patients at a certain health facility who received at least one injection, expressed as:

The vast majority of patients were interviewed in government health facilities, therefore the results should be considered representative of practices in the public sector only (Table 4). In Lebak, 71% of the patients interviewed received an injection. In Lombok, the mean injection rate in all health facilities was even higher: 87%. This difference between Lebak and Lombok is highly statistically significant (**p=0.001).

Table 3: Injection use in health facilities - Prescription analysis - Uganda²² (Percentage of prescriptions containing at least one injection)

²²N=total number of prescriptions reviewed.

Table 4: Injection use in health facilities - Exit interviews - Indonesia²³ (Percentage of patients who received at least one injection)

²³N=all patients interviewed at health facilities. In Lebak, five patients were interviewed in private practices of nurses and midwifes. Because of their small number, this data is excluded here.