How to Perform ISS and ISE Analysis Using SAS

For outputs from individual studies to be considered sufficient, they must fully address regulatory questions on their own, without requiring readers to combine information from multiple tables. Integrated Summary of Safety (ISS) and Integrated Summary of Efficacy (ISE) analyses are crucial components in regulatory submissions for clinical trials. ISS/ISE analyses using SAS involves several steps. Here's a general overview of how you can conduct these analyses:

Data Preparation and Exploration

Before starting ISS and ISE analyses, it’s important to make sure your clinical trial data is in a standardised format, like the Study Data Tabulation Model (SDTM). Using standardised data makes it easier to integrate and analyse data from different studies.

The first step in data preparation is merging individual study datasets into a single integrated dataset. This process involves aligning records based on a common key, such as subject_id, to create a comprehensive view of the data.

/* Importing individual study data */
proc import datafile="path_to_study_data.csv"
out=study_data dbms=csv replace;
getnames=yes;
run;

/* Merging individual study datasets */
data integrated_data;
merge study1(rename=(ae=adverse_event))
         study2
         study3;
   by subject_id;
run;



Once your data is integrated, the next step is to define key analysis variables and endpoints for safety and efficacy assessments. This includes identifying variables like adverse events (AEs) for safety and efficacy assessments Use SAS procedures to explore the dataset, ensuring all necessary variables are present and correctly formatted.

/* Exploring the dataset */
proc contents data=integrated_data;
run;

/* Generating descriptive statistics */
proc means data=integrated_data n mean std min max;
   var age weight height;
run;

Safety and Efficacy Analysis

Safety analysis focuses on identifying and summarising adverse events (AEs), serious adverse events (SAEs), and other relevant safety data. These analyses often require subgroup analyses to evaluate safety outcomes across different patient demographics or treatment groups. Next, define efficacy endpoints such as primary and secondary efficacy variables, response rates, and time-to-event outcomes. Generate safety tables and listings that summarize AE data by preferred term, severity, and relationship to treatment. Additionally, create efficacy tables and figures to highlight key efficacy findings, including treatment comparisons and confidence intervals. Finally, conduct efficacy analyses using SAS procedures that match the study design, and perform subgroup and sensitivity analyses to examine the impact of patient characteristics or protocol deviations on efficacy outcomes.

/* Generating frequency tables for adverse events */
proc freq data=integrated_data;
    tables treatment*adverse_event / nocum nocol nopercent missing;
    where phase='Phase 3';
run;

/* Calculating summary statistics for efficacy endpoints */
proc summary data=integrated_data nway;
    class treatment;
    var efficacy_endpoint;
    output out=efficacy_summary mean= std= min= max= /autoname;
run;

Integrated Analysis

After conducting safety and efficacy analyses, the next step is to integrate these results to evaluate the overall benefit-risk profile of the investigational product. This integrated analysis combines safety and efficacy data to produce a comprehensive summary. Finally, conduct sensitivity analyses to assess the robustness of the conclusions derived from the integrated analysis.

/* Combining safety and efficacy data */
data combined_data;
    merge safety_data(in=a) efficacy_data(in=b);
    by subject_id;
    if a and b;
run;

/* Generating integrated summary tables */
proc tabulate data=combined_data;
    class treatment;
    var adverse_event efficacy_endpoint;
    table treatment,
        (adverse_event efficacy_endpoint)*mean;
run;

Reporting

The final step in ISS and ISE analysis is preparing comprehensive reports that summarise the results. These reports should include descriptive statistics, graphical presentations, and statistical summaries, all formatted according to regulatory guidelines, such as CDISC SDTM standards.

/* Exporting results to a PDF file */
ods pdf file="ISS_ISE_Report.pdf";
proc report data=combined_data nowd;
    column treatment adverse_event efficacy_endpoint;
    define treatment / group 'Treatment Group';
    define adverse_event / analysis mean 'Mean AE Rate';
    define efficacy_endpoint / analysis mean 'Mean Efficacy';
run;

ods pdf close;

ISS and ISE analyses are critical for a successful drug approval. From the initial data preparation to the final reporting, each step plays a crucial role in ensuring that your integrated summaries are both accurate and comprehensive. This article provides a foundational example for conducting ISS and ISE analyses using SAS. By applying the principles and methods outlined here, you can develop well-structured and insightful analyses that align with regulatory requirements. Following these basic steps can improve the quality of your submissions and support more effective interactions with regulatory authorities, contributing to the successful approval of new therapies. At Quanticate our statistical programmers provide ISS ISE services and can support your regulatory submission. Contact us for more information.