Survey Says: Get the Data Right

May 1, 2009

Applied Clinical Trials

Applied Clinical Trials, Applied Clinical Trials-05-01-2009, Volume 0, Issue 0

With so much depending on accurate data, taking a do-it-yourself approach to scoring can be risky.

What's worse than spending millions of dollars on a clinical trial only to discover your findings are wrong? Learning too late that those errors could have been prevented. Health survey administration errors, data entry mistakes, and mathematical miscues are just a few of the common missteps during clinical trials that have cost companies millions and delayed the introduction of new drugs. In one well-known population study, for example, the first question of the survey was inadvertently left off when the questionnaires were reprinted.

Mark Kosinski

In another case, the questionnaire responses in a large clinical trial were mislabeled, leading to significant errors in scoring. Also, too often, companies try to write their own survey scoring programs.

Even companies with experience and expertise find the task of coding and reproducing scoring algorithms extremely challenging. With so much depending on accurate data, taking such a do-it-yourself approach can be very risky.

What it means

When data and scoring errors occur, the impact can be far reaching. You could miss out on a huge opportunity if bad data masks a benefit of your drug that could open up a new, unexpected market.

Fortunately, there are tools available that can help protect the time and money that you invest in health surveys.

Software exists that standardizes the scoring and processing of data and tests for data entry and coding errors. This software can simplify the process and provide reliable results you can trust.

Since the quality of your data is so important to your success, look for a scoring solution that can evaluate data quality and alert you to any data problems that exist.

Since humans are administering and completing the surveys, there are a number of things that can go wrong during the survey process. Questions can be skipped or answered incorrectly. The survey might be administered improperly or survey responses could be keyed into a database incorrectly. Make sure the scoring software you eventually choose can recognize these types of problems so that they do not corrupt your results.

When a partially completed survey is discovered, some of today's scoring programs can actually help fill in the blanks. Before you purchase your scoring software, find out if it includes this feature. If so, you may be able to save substantial time and money in the long run.

Other considerations

When deciding on a scoring solution, ask yourself whether you want to install and run the software at your company or if you want a patient reported outcome (PRO) expert to handle it for you. This person can help interpret your data as well.

If you decide to go the PRO route and you want the convenience of submitting your data on the Web, choose a scoring solution that offers that option.

On the other hand, if you have data privacy and security concerns and you want to score your data within the confines of your own software system, make sure the scoring software you're thinking about purchasing can be easily integrated into your own system.

Health surveys provide invaluable insight and information throughout the clinical trial continuum. Therefore, be sure the information you collect tells the right story. This is achieved by making wise choices about how you are going to score and analyze your data.

Mark Kosinski, MA Vice President and Senior Scientist QualityMetric Incorporated mkosinski@qualitymetric.com

Related Content:

A Closing Thought | CRO/Sponsor