Page 55 - PATIENT REGISTRY DATA FOR RESEARCH: A Basic Practical Guide

should be made transparent to the research team and the key stakeholders and also be clearly

documented in Statistical Analysis Plan (SAP).

Since the generation of output from these patient registry reports is regarded as a

repetitive work, therefore it is necessary to select an appropriate statistical software. This

software should incorporate the use of programming codes in the analysis. The recommended

statistical software is STATA (Copyright 1996–2018 StataCorp LLC), SAS (Copyright ©

2011, SAS Institute Inc., Cary, NC, USA) and R (R Foundation for Statistical Computing,

Vienna, Austria). Besides that, it is also necessary to ensure that the computer has fulfilled

certain competency requirements, such as the requisite standards for a sufficient RAM and

also appropriate working processor, in order to ensure that it has the capacity to conduct

statistical analysis on an extremely large data set.

If the analysis is not generated by using the programming codes, then it is important

to carefully document all the necessary input processes and outputs obtained from the

analysis in a structured work diary for our future reference. Regardless of whether the

analysis has been generated by programming codes or not, all the input processes (for

example: handling duplicates, matching and combine datasets, setting the conditions for

analysis, data cleaning and management, handling missing values and developing an overall

framework for the analysis) should be fully documented in the Statistical Analysis Plan,

which shall serve as a useful future reference.