3.7. Patient Pathway Analysis

The Patient Pathway Analysis (PPA) methodology describes the alignment between patient care-seeking and TB service availability. The results can inform programmatic priority setting and planning for more patient-centered availability of services.

The output of the PPA displays five core metrics to determine whether public health programs and service availability align with patient care-seeking preferences:

• Place of Initial Care Seeking: Proportion of patients that initiate their care-seeking journey at each sector and level of the health system.
• Diagnostic Coverage: Proportion of health facilities with a given TB diagnostic service, by facility level and sector (informal, public, private).
• Access to Diagnosis: Estimated proportion of care seekers that access diagnostic services on their first visit.
• Treatment Coverage: Proportion of health facilities that have anti TB medicines in stock and that can supervise patients during the treatment initiation or continuation phases.
• Access to Treatment at Initial Care Seeking: Estimated proportion of care seekers that access facilities with TB treatment services on their first visit.

1. Get familiar with the PPA process

A comprehensive step-by-step guide on how to conduct a PPA can be found in the PCF learning Management System (LMS), see the e-Learning section for more details. The videos below provide an introduction and overview.

2. Collate the data

The essential data sources for each metric are shown below. However, a complete overview of data sources per PPA metric can be found in the patient pathway guide.

If the Demographic Health Survey (DHS) data will be used, ensure registration with the DHS is done one month (at a minimum) before the process starts. This request can be submitted here: The DHS Program – Data.

3. Clean the data

Data from the different data sources should be cleaned to ensure data is quality. This is the process of finding and correcting untrustworthy, erroneous, or outdated information from a database It aids in the detection of data that is missing, wrong, inaccurate, or irrelevant. Data cleaning can be done by monitoring errors, validating data accuracy, scrubbing for duplicate data, conducting quick analysis on the data, and communication with the team.

The quality of the data also impacts successful data upload into the PPA wizard and the quality of the PPA outputs (results). Data cleaning can be done using any statistical software package, e.g., MS Excel or STATA, the PPA.

4. Upload cleaned data into the PPA Wizard

The PPA Wizard How-to-Guide (click button above) provides a step-by-step guide on how to register on the website, upload data (with a description of the different data types and formats required), and finally generate the PPA results. High-level overview of the steps include:

Select PPA Metrics: number of health facilities and place of initial care seeking are pre-defined, TB service capacities can be added and defined as desired (up to 4 different diagnostic and treatment services)

Provide Data: Upload the data source per PPA metric, one source can be used multiple times

Identify Variables: Select the columns containing values for facility type, health sector, and health service for each dataset. Select columns containing geographies if you’re doing a subnational PPA.

Map Facilities: Assign level and ownership to each facility. Ensure the variables are named to match the respective datasets as this makes the process more efficient.

Map Geographies: Assign geography (or hierarchy) to the facilities (where available).

Create Output: Clicking this button creates the output files including the data sets and the visuals referred to as the PPA results.

5. Interpret the PPA Wizard output

This step involves the discussion of the results. To guide the discussions,

• Identify gaps between care-seeking and health facilities at each level
• Identify gaps between care-seeking and diagnostic services
• Identify gaps between care-seeking and access to diagnostic services
• Identify gaps between care-seeking and treatment services

Example of a PPA Result (output) generated by the PPA Wizard

6. Conduct sensitivity analysis

A sensitivity analysis determines how different values of an independent variable affect a dependent variable under a given set of assumptions. For this process, conduct a sensitivity analysis using different datasets, while acknowledging the limitations of each dataset such as the target group, year conducted, etc. Based on risk groups in the country (e.g., smokers, low socio-economic status (SES), men, elderly), conduct stratified analysis for each group.