Lincoln Medical Center Patient Satisfaction Survey 2014
(Drs. Aliu, Alarakhia, Biscotti, Grice, Habing, Wall, Laverdiere, Ireland).
Patient satisfaction surveys are at once both the easiest QIP (quality improvement project) to get off the ground, and the most difficult. What’s easier than handing out a survey? Well, think about it.
The patient satisfaction survey is a way to break out of your day to day rut, a way to start stepping back to appreciate the effect of your interaction with your patient. Physicians are used to measuring these effects indirectly, vaguely, intuitively. We measure the blood pressure on the next visit and see if it’s gone down. We ask how many cigarettes they’re smoking now. We hope for a change of countenance, a sparkle in the eye, a reported improvement.
An action research stance in medical practice is about identifying a problem, researching, measuring somehow the current state of affairs, doing your intervention, and then measuring again. The wheel doesn’t stop turning: you use these results to jigger your intervention some more, and try it again, hoping for continual improvement. Really, this is the crux of the quality improvement project.
The measurement tool is concern number one. Too difficult, and you’ll get blanks. Too flimsy, and you’ll get no valuable data. Has the tool been validated? Does it measure what you think it measures? Is it reliable?
Validity speaks to the measuring device, the instrument itself. A valid instrument measures what it’s meant to measure. A metre stick measures length, not volume. Reliability speaks to consistency, whether the instrument will work time after time (McMillan, 2012). A valid test is said to be always reliable.
What’s more important? The question of whether validity or reliability is more important to research results is essentially moot. A measuring instrument that measures the wrong thing will give no useful data. If this invalid instrument would very reliably measure the wrong thing time and again, what use is that? However, if the valid instrument is in itself unreliable, the study cannot be completed. True, with a questionnaire type instrument, it would be difficult to understand how it would be valid but not reliable: perhaps a questionnaire written on the sand! However, in scientific inquiry, a valid measuring technical instrument could certainly break down, and need calibration, or repair. One must be able to rely on the instrument to perform as one observes data. Unmeasured, undocumented events have no place in scientific inquiry.
There are a lot of standard measuring devices out there. Included in these are cognitive, affective, and projective instruments: all can have a role in your research question. Cognitive instruments are tests, aptitude tests. Affective instruments measure attitude, values, an example being a personality inventory, or the HAMD7 score on your depressed patient. Projective tests, such as the inkblot, measure responses to ambiguous stimuli.
We have decided at our clinic to tweak a previously validated patient satisfaction survey for our own use. At the heart of the appended survey is the 2013 UK General Medical Council patient feedback questionnaire. This original document was felt to be a bit too long when it was used on paper and clipboards in 2013.
The current clinic version is cleaner, simpler, involves visual Likert scales, is presented glowingly on an iPad, and has only one page! The improvement is not only visual moving into the digital arena: Doug Kavanagh from CognisantMD (for PS Suite EMR) had a couple of surprises for us. This template not only collects information completely confidentially, but when all these results downloaded and analyzed in aggregate, changes can be measured from year to year, for each patient!
If this wasn’t clear, let me rephrase this concept. This template questionnaire is completely confidential, and anonymous. No health care practitioner can ever find out what patient gave what result. However, when this survey is run again, six to twelve months from now, not only can a general average response be calculated (ie an overall change in approval ratings, etc.), changes in response can also be measured per patient.
Now that’s something a paper questionnaire will never give you.
Valid? I think, and hope, that this measuring instrument measures what it’s designed to measure. It’s fairly well based on a previously validated instrument. Let’s hope. Reliable? Well, this questionnaire hasn’t been “run off” on the photocopier. It’s got digital magic dancing between Likert scales. I suppose it could run off the rails.
But this digital alien wants it to work!
If you’re running PS Suite, and interested in something like this, email Doug Kavanagh at firstname.lastname@example.org.
McMillan, J. (2012) Educational Research: fundamentals for the consumer. Boston: Pearson
GMC Questionnaires and Resources. (2013). General Medical Council: Regulating doctors,
Ensuring good medical practice. Retrieved from
Our 2014 Patient Satisfaction Survey: