Are goniometers obsolete, especially in the coming era of telerehab?
Those of you have attended the Doctors Demystify flagship course, Current Science for Hand Therapists, know that I stress basing our professions on scientific data whenever possible and that my quick definition of science is an endeavor which is measurable and reproducible.
Joint motion falls into that realm. We measure it with goniometers, and various studies show that the measurements are reproducible with acceptable intra- and interobserver errors. Yet goniometers are awkward and are typically scaled at 5-degree intervals, meaning that a measurement is unlikely to be more accurate than +/- 5 degrees of the actual angle. Can we do better, especially in the burgeoning era of telehealth, where we cannot expect patients at home to have a goniometer or know how to use it?
Three papers within the past 4 years address this issue by testing the accuracy of smartphone photos to determine elbow, wrist, and finger motions. Another article, published a month ago, uses a unique smartphone app for measuring finger joint ROM.
First, the smartphone photos.
Investigators in Minnesota paired 32 subjects without known wrist pathology and had them take smartphone photos of each other’s wrists in standard positions. Prior to taking the photos, the participants watched a YouTube video developed especially for this study. The video demonstrated proper patient and limb positioning and camera placement. Once each pair of subjects had photographed each others’ hands, two investigators also photographed the subjects’ hands and then measured the subjects’ wrist angles with a goniometer. Results: all measurements were within 1-7 degrees of each other, and none of the differences were statistically significant.
Some of the same investigators used identical methodology to compare smartphone photographs with standard goniometry of elbow flexion and extension. Again, a YouTube video explained how to the subject should stand and position the shoulder, elbow, and wrist and how the camera should be positioned. Thirty-two subjects participated. Results: the average range of motion was zero to 129 degrees of motion for both the photographic and goniometer measurements. Interobserver error was statistically insignificant.
Researchers in Massachusetts studied 50 patients with a total of 94 digital joints contracted more than 20 degrees from Dupuytren disease. The patients were given a handout that described the proper positioning of the hand and camera. The first set of photographs was taken by the patient’s family member or friend to simulate a home setting (“untrained photographer”). A second set of photographs was taken by one of the investigators (“trained photographer”) One of the investigators, a fellowship-trained hand surgeon, then performed manual goniometry. Results: the mean contracture for al joints was 39 degrees for manual goniometry, 35 degrees for trained photography, and 36 degrees for untrained photography. The differences were not significantly different statistically.
Now for the cutting-edge technology:
Investigators from Japan have developed a smartphone app that uses the built-in accelerometers (the same ones that track your walking/jogging steps in the fitness apps) to determine how far off horizontal the phone is when it is placed on the dorsum of a finger. The fit is better, especially on interior fingers, when a ruler is taped to the phone as seen in the photo. Once the movement plane with respect to the ground is established, the investigator places the smartphone (or its extension) on the dorsal midline of the distal phalanx and presses the start button. In 3 seconds, a bell chimes, and the phone is moved to the dorsum of the middle phalanx. Then at 3 second intervals, the phone is moved progressively more proximally next over the proximal phalanx and finally over the metacarpal.
The investigators then measured 252 joints of 63 fingers of 20 patients, both active/passive and flexion/extension for a total of 1008 values. They compared the smartphone results with goniometer measurements. The interclass correlation coefficient was .927 with 95% confidence interval of .902-.944, in other words, the smartphone accurately measured the angles. That is not surprising (otherwise they probably would have not published the paper). What is even more helpful than knowing that the method is accurate is that it took only half the time that the goniometer measurements took.
COMMENT: Not all of our patients have smartphones, and the accelerometer-based app is presently only available in Japan and for Android operating systems, so it won’t work on iPhones; but both issues will likely dissolve quickly. Of course, the smartphone transmission of photos or digital measurements is not HIPAA compliant, again manageable as long as patients understand that their joint measurements might escape into cyberspace. (Who would be interested?) Overall, these papers offer bright notes in dark times to remotely obtain measurable and reproducible ROMs. I will now be able to put my goniometer in the keepsake box with my slide rule and 35 mm camera.