Today, 8 November 2007, is World Usability Day.
World Usability Day was created in an attempt to ensure that services and products important to life are easier to access and simpler to use.
This year's theme is Healthcare. There are events happening across the globe, from Bulgaria to Perú. We thought we would add to the occasion by taking a look at the usability of a healthcare form, namely the questionnaire given to new patients.
New patient questionnaires are extremely common in the world of healthcare, at least in Australia. My local doctor has one, as does my dentist and my podiatrist. The information these forms collect is the basis for the patient's file and provides the starting point for the provision of services.
Because new patient forms are so prevalent, and because they frequently ask the same questions, it would be reasonable to assume that most people have very little trouble filling them out. However, an analysis conducted by Formulate showed that unfortunately, this is not the case.
Analysing completed forms
We looked at a random sample of 60 new patient questionnaires that had been filled out by people visiting a local physiotherapist. Some of these forms had been filled out recently, others were completed by patients who first visited the practice a long time ago. The patients themselves came from all walks of life and were of all ages, from school students to retirees.
The questionnaires were all paper-based and were completed when the patient first arrived at the practice. Each questionnaire was a single page of just over 30 questions.
Looking at filled-out forms won't pick up all the issues they contain, but is a good way to get a feel for some of the more obvious problems. For example, it will reveal places on the form where respondents aren't given enough room to write their answers.
Problems were prevalent
Reviewing the 60 randomly selected forms this way, we were surprised to discover that every single form contained at least one error.
Half of the patients made between 1 and 5 errors on their form, with the other half making 6 or more errors. The maximum number of errors made on a single form was 11.
There were 85 different types of errors. The most common error was failing to answer a question that was relevant to the respondent.
What sort of errors were they?
The errors patients made varied in seriousness.
For example, a less serious — but nonetheless interesting — error related to the question about being pregnant. This question had three answer options: "yes", "no" and "not applicable". The analysis found that in 15 cases, the male form-filler gave the answer "no", rather than "not applicable", to this question.
This pregnancy question example is obviously not a critical one. However even it would need to be corrected at the data entry stage, something that would cost the practice time (more on this later in the article).
In other cases, however, problems with the form have more serious implications. Consider the question shown in Figure 1. There are a number of problems with the way this question is worded, and these problems could lead to the practitioner not having the right or enough information about the patient. In particular, some key conditions that are relevant to the treatments provided by a physiotherapist, such as anxiety, are not included.
The question is also quite ambiguous. Should the form-filler report only conditions for which they are currently experiencing symptoms? What if they have had one of conditions in the past, or experience symptoms only in certain circumstances? What if the condition is mild or hasn't been formally diagnosed?
Is it terminal, Doctor?
To determine just how serious the problems with the form were, we categorised the errors according to what needed to be done to correct them:
- Errors that could only be corrected through discussion with the patient
- Errors that could be corrected through discussion with the patient or by consulting another source, such as a phone number directory
- Errors that could be corrected by the person doing data entry, without consulting any other sources.
We then estimated how long correcting each type of error would take.
We found that the time required to address all the issues in the 60 forms would be approximately 110 minutes, or 1.8 hours.
Whilst 1.8 hours may not seem like much time, imagine how much more time would be required if we were talking about a more complex or less common form.
Moreover, poor form design such as this causes significant frustration for the people who have to fill the form in. In a medical services situation, the individual doesn't always have a choice of provider. But in other situations, such frustrations may well lead the form-filler to choose your competition instead of you.
Hopefully this article demonstrates that a form which appears to be working well can in fact be plagued with problems. These problems can cost the owning organisation money and time. More importantly, they can lead to lower quality service delivery or even put someone's life at risk.
The story is particularly grim if you consider that the analysis discussed in this article only involved reviewing completed forms. There was no checking of answers against external sources, no observation of patients filling out the form and no examination of the relationship between the form and the system into which the answers was manually entered. Had we performed these sorts of analyses, no doubt more problems would have been uncovered.
So on this, World Usability Day, why not stop and have a think about the forms your organisation has. Is it possible your forms aren't working as well as you might think? Would problems with your forms have serious implications for you as an organisation, through inappropriate service or product delivery, causing harm to others, or reduced user satisfaction? If you answered yes to either of these questions, you might want to consider talking to a professional form designer.