URL: , Publisher:  Chicago Clinical Practice Enhancement and Anesthesia Research Foundation , Webmaster:  Erik Szabados , David J. Mingay, Ph.D. Department of Anesthesia and Critical Care University of Chicago, Chicago, Illinois Current position: Statistical Research Division US Census Bureau, Washington, DC

Many patients and physicians use the Internet to obtain medical information. Administering a preoperative questionnaire to patients via  the Internet before a preoperative clinic visit offers substantial cost savings without compromising quality of care. For example, if no medical problems necessitating further investigation are identified, a clinic visit can be canceled unless the patient wants it. If a visit is needed, it may be shorter and necessitate less clinical staff time because much of the patient’s medical history has already been obtained. HealthQuiz is a preoperative assessment questionnaire that originally was administered on a dedicated small portable computer 1and is now available on the Internet without charge and by touch-tone telephone for the cost of the toll-free call.

Internet users have little patience for complex, slow, or unappealing Web sites. 2Even experienced users may have taken few, if any, questionnaires on the Internet. Thus, to minimize break-offs from the site and to reduce errors in responding, the questionnaire must be easy and satisfying to take. I evaluated the user-friendliness of the HealthQuiz Internet questionnaire. To inform the evaluation, after approval from the institutional review board of the University of Chicago Medical Center, Chicago, Illinois, I observed 12 patients with varying levels of Internet experience taking the questionnaire and administered questions about the task.

Questionnaire Design and Administration

Each patient was given a sheet with brief instructions, the Web site’s universal resource locator (URL), and his or her identification number. After entering the URL and identification number, the patient clicked on buttons to select the interface and begin the interview. The first questions ask for the patient’s gender, age, race, height, weight, blood pressure, and invasiveness of the surgery. An average of 124 health questions with the answer categories of yes, no, and not sure are then administered.

There are two interfaces, termed textual and graphical. The textual interface loads relatively quickly, has several questions on each page that are answered using small radio buttons, and is suitable for slower Internet connections. The graphical interface has one question per page, has larger text and answer buttons, and is most suitable for a faster connection. *

Textual Interface

The method for entering height in inches on the textual interface was confusing. The box on the left had the numbers 0 and 1, and the box on the right had 0–9. This violates the Internet convention of using a single drop-down box for low double-digit numbers. 2The layout of the answer buttons for the health questions also caused confusion. The word yes was equal distance from the buttons for yes and no, and the word no was equal distance from the no and not sure buttons. Four subjects clicked the wrong answer button at least once. Finally, it was not readily apparent how to return to the previous question because there was no button on the screen with this function. Four patients did not realize that they could use the back button on the browser for this purpose.

Graphical Interface

The relatively simple design of this interface made taking the questionnaire easier than with the textual interface. For example, most patients found it easier to respond by pressing buttons or typing in numbers than clicking on drop-down boxes. For the health questions, the size and prominence of the text and the response buttons made answering easy. The large back button clearly indicated how to return to a previous question. However, the design of the progress bar caused confusion and annoyance because it reset to zero after the demographic questions had been asked. All patients had assumed that the bar showed their progress throughout the entire questionnaire rather than a part of it.

Changes Made

Changes were made to the Web site to resolve the difficulties identified. On the textual interface, the two drop-down boxes for height in inches were replaced with a single drop-down box. The response buttons for the health questions were placed vertically to reduce the likelihood of selecting the wrong button. On the graphical interface, the progress bar for the demographic questions was removed so that patients see a single non-resetting bar. A small-scale follow-up study suggested that the changes have significantly improved the ease of using both interfaces.


Internet questionnaires must be easy and satisfying to take. Evaluation of the HealthQuiz preoperative questionnaire identified several problems with both interfaces that caused patients difficulty and irritation. Changes were made to the interfaces to resolve these problems.

Roizen MF, Coalson D, Hayward SA, Schmittner J, Thisted RA, Apfelbaum JL, Stocking CB, Cassel CK, Pompei P, Ford DE, Steinberg EP: Can patients use an automated questionnaire to define their current health status? Med Care 1992; 30: MS74–84
Nielsen J: Designing Web Usability. Indianapolis, New Riders Publishers, 1999