Statistical programmers

It is best to begin by describing the environment in which a statistical programmer works in the pharmaceutical industry. Then we explore the fundamental principles that should guide you in your day-to-day work. These principles permeate all of the tasks that you do on a daily basis and, if kept in mind, they will keep you from going astray in your statistical programming duties. 

Regulatory authorities govern and direct much of the work of the statistical programmer in the pharmaceutical industry. It is important for you to know about the following regulations, guidance, and standards organizations. 

CFR - Part 77 is a federal law that regulates the submission of electronic records and electronic signatures to the FDA. Of particular interest to the statistical programmer are the following requirements of Part 11 ... 

Within any pharmaceutical company or contract research organization, there are groups and individuals outside the statistics department that you work with. Fet s take a look at the functional groups a statistical programmer interacts with most. 

In an optimal world, the CRF is perfectly designed to answer the questions of the study and the clinical data management group will have cleaned the data to perfection. However, to be a good statistical programmer in the clinical trial arena, you must always keep a lookout for errant data and program defensively. Defensive programming lets you account for all possible clinical data permutations. 

Clinical trial data come to the statistical programmer in two basic forms numeric variables and character string (text) variables. With this in mind, there are two considerations for all numeric and text variables. All data should be cleaned if they are needed for analyses, and any data entered asfree-text variables should be coded or categorized if they are needed for analyses. 

Before the statistical programmer receives data that are ready for analysis, the clinical data management group cleans the data. This is done through a query process, which is built into the clinical data management system. The clinical data management query process usually looks like this ... 

Clinical trial data come in two basic forms numeric variables and text variables. Numeric variables are easy for the statistical programmer to handle. Numbers can be analyzed with SAS in a continuous or categorical fashion without much effort. If a numeric variable needs categorization, it is easy enough to categorize the data within SAS. For example, if you had to classify patient age, a simple DATA step such as the following might serve well. 

The problem for the statistical programmer in categorizing data comes from text variables, or more specifically, free-text variables. A free-text variable is one that may contain any characters and is typically limited only in length. As an example, let s say you need to summarize the adverse events for a set of patients in a trial. The following SAS code shows the data and a quick summarization of the adverse events. 

An essential detail for the statistical programmer to watch for in prior or concomitant medications data is whether or not the start and stop dates are important for analyses. Unfortunately, it is often the case that the importance of the timing of prior or concomitant medications is not determined until after much of the data have been entered or even after the database is closed to entry. For instance, it may be decided later that a specific concomitant medication has to be watched carefully for interaction with a medication used in the study. If insufficient attention was placed on the quality of the medication start and stop dates, then determining whether there is overlap with study medication is difficult if not impossible. 

Laboratory data can pose a challenge to the statistical programmer in many ways. 

The adverse event form is a cornerstone of patient safety monitoring, and as such it contains very important data. There are several data issues for the statistical programmer to be concerned about here. 

Other data sets may be found within the IVRS system that prove useful to the statistical programmer as well. Often the IVRS collects several baseline patient characteristics that are used in the stratification of the randomization scheme and subsequent assignment of study therapy. Finally, the preceding examples show in detail what the treatment variable is, in the treatment column. It is more often the case that the treatment variable is coded, such as A or B or C. It is of paramount importance that you know with absolute certainty how the treatment code can be properly interpreted. 

XML will become more integral to the work of statistical programmers in the pharmaceutical industry as the standards, applications providers, and vendors make more use of this technology. Eventually you should expect the FDA to move away from SAS transport files to XML files as their standard data format for electronic data submission. 

In this section we take the aforementioned principles and guidelines for analysis data sets and apply them to creating the most common analysis data sets. The critical variables, change-from-baseline, and time-to-event data sets are presented. Although these are the most common analysis data sets that a statistical programmer will encounter, they are by no means all of the possible analysis data sets. When it comes to analysis data sets, there is no limit to the diversity of data that you may have to create. 

Present output This step involves presentation of the summarized data. There is a wealth of options available here for the statistical programmer. 

Commercial software (e.g. word processing, database and statistical programmes) may be considered to be sufficiently validated. However, there may be a need to validate laboratory software configuration/modilications. 

---