Data on referral source provides insight to the evolution of a program’s case finding and indirectly assesses the performance of the program’s outreach workers. In general, a program that has good community mobilisation and awareness raising, it can be expected that source of referrals will start out as predominantly referrals coming from outreach workers and then slowly evolve into more self-referrals later on in programme implementation.

Data requirements

Data on referral source can be collected from the outpatient care treatment cards and recorded using a tally sheet.

Analysis of data

A Pareto chart is the ideal graphical approach to analysing referral source data. A Pareto chart is a specialised bar chart that presents factors or causes of a problem or issue in decreasing severity. A similar approach using a tally sheet to collect and tabulate data as with the time-to-default or admission MUAC can be applied to creating a Pareto chart. The first column of the tally sheet will be for the list of referral sources while the second column will be filled out with the respective tally of each referral source based on the beneficiary records. However, to create a Pareto chart, an additional step of sorting the referral sources by order of their respective tally from the most common to the least common is needed. Once sorted, the tally sheet can be re-created with the most common referral source at the bottom of the list and the least common referral source at the top of the list and then the corresponding hash counts entered into the second column. The Pareto chart can be created using a spreadsheet or a statistical package by creating a bar chart from the sorted data (from most common to least common) of referral sources. Figure 1 is a Pareto chart created using a spreadsheet.

Figure 1: Referral sources of a CMAM program in Western Kenya

Data courtesy of World Vision Kenya


If the program is getting majority of its referrals from volunteers, this can be taken as an indicator of high outreach workers activity. On the other hand, if the programme is admitting more self-referrals, this can be understood as the awareness amongst the community regarding the programme is so widespread that mothers or carers are coming to the clinics on their own to access services. It should be noted that an increase in self-referrals in comparison to outreach workers’ referrals should not be taken to mean as a decrease in outreach workers’ activity. In fact an increase in self-referrals, which is said to be by-product of the increase in awareness of mothers and carers, can be seen as a result of the community sensitisation and mobilisation activities done by active outreach workers. This would mean that data on referral source should be analysed in the context of the current phase of the program being investigated and should be triangulated to other measures of outreach workers’ performance and level of activity.

The program described by Figure 1 above is already a well-established program that has been running for more than 2 years and over this period have consistently shown good outreach workers’ performance with regard to case-finding. Hence, the pattern of referrals shown in Figure 1 reflects a strong level of outreach workers’ activity and a general increased awareness of the program by the mothers and carers.

Download method guide