Types of Techniques

There are three computer techniques that are useful in quality control programs. The first is data base management or "electronic filing." Data base management includes the creation and updating of computer-based records. These records may be kept solely for quality control purposes, or they may be made accessible for other functions, such as billing. The establishment of an effective data base management system is the first step in establishing a computerized quality control system.

The second technique is the use of statistical tools to analyze quality control data. These tools are necessary to identify legally significant events that may be obscured by the mass of legally insignificant data. There are two types of statistical tools that are useful for analyzing quality control data. The most commonly used tools are descriptive statistics. As the name implies, these tools are used to describe the data, providing information such as averages, means, and so on. The second group of statistical tools involves techniques that measure trends through time.

Trend analysis techniques are important for quality control because they allow the detection of potentially significant events before they result in legal liability. An example of this type of event would be an increasing number of surgical wound infections. While absolute number of infections might still be within allowable limits, a continued increase would push the number of infections into the legally significant area. Such an increase might be obscured by the use of descriptive statistics, but it would be detected by trend analysis.

The third set of techniques involves computer aids to decision making. These techniques range from simple computerized checklists to complex artificial intelligence programs. The use of computerized checklists is a valuable aid to carrying out preventive law checkups. While checklists are usually too limited for analyzing complex or rapidly changing information, the use of a computer allows the use of branching checklists. These checklists can be made very comprehensive while still remaining easy for unsophisticated personnel to use.

The use of artificial intelligence techniques is still in the experimental stage. A common artificial intelligence technique uses computer programs that learn from experience. These programs can survey information for simple patterns, then use the patterns to aid in the detection of more complex patterns. From our experience with these techniques, we believe that their greatest value will be in managing problems such as infection control and medication errors.