One of the most common tools used to measure quality improvement wherever services are offered is the survey. It has many good applications if it is well designed and can provide a lot of information if it is properly analyzed.

When and why should you use surveys? Of course one of the most frequent uses is with patients and clients. In fact, Medicare is beginning the H-CAHPS survey of hospitals to find patients views on many different facets of their service. The survey is voluntary and results will be posted online in 2007. It is becoming a necessity for physicians and healthcare professionals to find out what their patients think as more and more information about care from insurance providers and other sources is being posted online.

The employees of any healthcare facility should be surveyed too. They should, after all, be considered clients. Since they provide many different services and are the front line connection to patients, their input about their work environment and satisfaction with it are to be taken seriously. You should not just be surveying doctors and nurses, but also maintenance, administrative, and all other groups at your site.

What are some of the basic ideas you should keep in mind when designing a survey?

1. The best questions are the anchored endpoint type. One end would be very satisfied and the other very dissatisfied or something similar, with a 7 or 10 point scale from one end to the other. The five point scale is too short. It doesnt provide enough variance.

2. You may ask a few yes and no or gather demographic data, as age.

3. Limit yourself to one open-ended question. They are too hard to provide statistically significant data.

4. When writing the questions, work with a team of representatives of the people who will be surveyed. This helps avoid bias and makes sure you have good questions.

5. Be sure to choose a random sample to survey. It is better to survey 30 or 40 randomly chosen patients or clients whom you interview or have fill out a survey while in the office rather than getting more surveys by subjects who return mailed surveys or volunteer in some other fashion. Voluntary surveys are basically worthless from a statistical point of view.

Once you have collected your surveys, it is time for analysis. For simple analysis, I suggest using Microsofts Excel. It has some dynamite graphics. Too, you need to get the mean and standard deviation of each question where appropriate. Remember, you want your mean to be as good as possible with low standard deviation. If you want some more insight and powers of prediction of what action you should take based upon the survey, I suggest you employ someone who is trained in statistical analysis or a statistician. The return on the investment of a professional analyst should easily exceed the expense.

Summarily, surveys are a valuable instrument in finding out the opinions of your patients and employees. Care should be taken in designing the instrument to insure there is no bias and that there is randomness in conducting the survey. Investment in professional analysis is well worth it. The result will be healthier patients who are more loyal, a definite financial plus.