Hi all,
We are wanting to use Qualtrics to track our patients metrics over time, and I know our situation means we have to think a little outside the box on how to set-up Qualtrics for our use.
Simplified, we have a longitudinal survey-style, wherein every patient completes 4 surveys at various time points (pre-treatment, post-treatment, 1-week, 1-month, 2-month, 3-month, 6-month and yearly). The 4 surveys will vary based on the patient's diagnosis (i.e. a depression patient will have a different batch of surveys than a stroke patient).
Ideally, we can analyse the data for an individual patient across time, but also look at the cohort's response to a survey over time. Secondly, some surveys will have responses from individuals with a variety of conditions (i.e. one survey will have responses from both depression and stroke patients) and it would be great to analyse one conditions response (i.e. look at all of the stroke patients data and exclude the depression patients).
We are unsure whether the best method is to set up all of our surveys that are triggered at various time points, or if we should create surveys for each separate time point (i.e. Beck's 1-week and Beck's 1-month).
I haven't had much support from our Account Director and it would be fantastic if someone from the community could help support us make the most of this platform!
How to best set-up Qualtrics for our healthcare practice.
+30
Hi isabellayoung ,
Before you build any surveys I think the most important thing would be to assess what kind of data you have in your patient information system, and identify what can be carried into your email list and surveys as embedded data. That will go a long way to being able to determine how to set up your interval surveys and how to organize the types of questions. If you are able to carry carry background data into your email list then you can use it to identify the different types of patients in your data.
You'll want to keep the number of different surveys you're managing as low as possible, within the limits of the types of questions you're asking. If the questions are the same at each interval you may be able to the same one for each point in time.
Finally, could any patients have more than one diagnosis? If so, that may indicate that you need to have a single survey that has questions for each diagnosis in different blocks. If not, you could keep each diagnosis in a separate survey.
Our patient information system has both personal information we could use as embedded data (email), but also patient identification numbers.
The surveys we use are psychometric questionnaires (e.g. Beck's Depression Inventory, Generalised Anxiety Disorder Scale), that are typically Likert scale-type questions, wherein a value is given to each response. The final score is typically a sum of all of the response values.
Each patient would complete the same set of 3-4 surveys at multiple time points.
And yes, some patients could have more than one diagnosis.
An example workflow at the moment: A patient comes to us with depression and anxiety seeking treatment. They will complete a Well-Being Questionnaire, a Sleep Questionnaire, and two surveys that track their diagnosis symptoms (i.e. Beck's Depression Inventory and Generalised Anxiety Disorder 7-item Scale). They will complete all 4 of these questionnaires at various time points, and we would like to see how they track on each individual questionnaire over time, but also compare every depression and/or anxiety patient's responses on the surveys. This is simple for the Beck's Depression Inventory, wherein all responders would be provided that questionnaire because they have the diagnosis of depression, whereas the Well-Being Questionnaire is completed by all patients, and as such the ability to expand responders based on diagnosis would be ideal.
+30
Thanks for the additional info. Given that the intervals for the Sleep, Beck's and Anxiety surveys are different from one another, then it's probably best to keep them as separate survey projects in Qualtrics. But you could use Actions to at least automate the initial distribution of those surveys after they fill out the well-being Q, if that is applicable.
Within each of the 3-4 surveys, it sounds like the questions will be generally the same at each time point, so I would use the same survey for each interval; setting up each interval as its own survey would probably complicate the analysis. If there are questions that are unique to a time period, then Blocks or Display Logic can be used to show them when appropriate. You can use an ID number in Embedded data to link respondents within and between surveys, if needed.
Leave a Reply
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.