Using a Robot to Treat Non-specific Low Back Pain: Results From a Two-Arm, Single-Blinded, Randomized Controlled Trial

Abstract

Non-specific low back pain (NSLBP) affects many people and represents a high cost for health care. Manual pressure release of myofascial trigger points is used to treat NSLBP and is very effective but difficult to standardize since it is provided by different therapists, which also suffer musculoskeletal complications from this highly repetitive activity. A robot designed for this purpose may help in reducing these problems. Here, we present data from a two-arm, single-blinded, randomized controlled clinical trial evaluating the efficiency of a therapeutic massage robot (ADAMO) in reducing NSLBP (clinicaltrials.gov, registration number: NCT04882748). Forty-four patients were randomly distributed into the two arms of the study (robot vs. control). A physician filled the Oswestry disability index (ODI) before starting the treatment and at the end of it, in a blind fashion. In addition, patients filled a visual analogue scale (VAS) after each of the 10 treatment sessions. The ODI and the VAS were analyzed as the primary and secondary outcome measures. Both treatments (robot and control) resulted in a significantly lower ODI (p < 0.05). On the other hand, robot-treated patients significantly reduced their VAS levels (p = 0.0001) whereas control treatment did not reach statistical significance. Patients of both sexes obtained similar benefits from either treatment. Overweight patients (body mass index ≥ 25kg/m2) in the robot arm benefited more from the treatment (p = 0.008) than patients with normal weight. In conclusion, the ADAMO robot is, at least, as efficient as regular treatment in reducing low back pain, and may be more beneficial for specific patients, such as those with excessive weight.

Keywords: body mass index; low back pain; overweight; perceived pain; robot; therapeutic massage.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Marín-Méndez, Marín-Novoa, Jiménez-Marín, Isidoro-Garijo, Ramos-Martínez, Bobadilla, Mirpuri and Martínez.

Figures

Figure 1

Representative photographs of the ADAMO device showing the robotic arm with the terminal handpiece (A), and detail of the handpiece (B) indicating the location of the thermal camera (a), the air heater (b), the cross selection laser (c), the air nozzle (d), the distance sensor (e), and the 3D camera (f). The black tube carries compressed air.

Figure 2

Consort flow diagram of participants through the study.

Figure 3

Comparison of the ODI (A) and VAS (B) scales before and after the treatment in both arms of the study. ODI values are expressed as box plots, which represent the interquartile range with the median as a horizontal line. Whiskers encompass the maximum and minimum values of the population. *p < 0.05, **p < 0.01, compared to initial ODI. VAS values are represented as the median (Q1–Q3) for each time point. ***p < 0.001, compared to initial VAS.

Figure 4

Comparison of the ODI (A) and VAS (B) scales, before and after the treatment in both arms of the study, taking into consideration the sex of the patients. Values are expressed as box plots, which represent the interquartile range with the median as a horizontal line. Whiskers encompass the maximum and minimum values of the population. *p < 0.05, **p < 0.01, compared to initial test.

Figure 5

Comparison of the ODI (A) and VAS (B) scales, before and after the treatment in both arms of the study, taking into consideration the BMI of the patients. Values are expressed as box plots, which represent the interquartile range with the median as a horizontal line. Whiskers encompass the maximum and minimum values of the population. *p < 0.05, **p < 0.01, compared to initial test.