Data case study

What influences the success of a medical procedure?

An international donor sponsored a pilot programme focused on improving the provision and uptake of family planning services (IUD devices). A pilot study was held at 6 hospitals in the Western Cape and a further 2 hospitals acted as a control.
Two aspects needed to be understood. Namely,
did the programme change attitudes to family planning and were there any factors that could predict the success of the procedure.

APPROACH: The project needed a quantitative survey and data analysis.


A baseline survey was run amongst healthcare providers at the beginning of the project and again at the end. The survey measured attitudes, knowledge, training attendance and job satisfaction. Surveys were paper based.

The comparison included a full gap analysis to identify and quantify change in attitudes and confidence as well as the penetration of training modules achieved. Confidentiality of response was critically important, so a 2-envelope system was designed with survey results and informed consent in separate envelopes linked with a unique code. The envelopes were opened by separate teams. 


We received anonymized patient data over a 28 month period. The data was supplied in some 300 spreadsheets which needed to be cleaned and combined. The data was then analysed to test what factors could be associated with the success of the procedure. In addition we used logistic regression to see if a specific factor affected the odds of an outcome.


  • The client was able to track how many health care providers the training reached and to what extent it changed attitudes.
  • In addition we were able to identify what factors were associated with successful outcomes (e.g. age, hospital,etc…) as well as whether specific factors could reliably affect the odds of a successful outcome.
Scroll to Top