The trusted source for
healthcare information and
Warm Baths for Sore Backs
Abstract & Commentary
By Dónal P. O'Mathúna, PhD . Dr. O'Mathúna is Senior Lecturer in Ethics, Decision- Making & Evidence, School of Nursing, Dublin City University, Ireland; he reports no financial relationship to this field of study.
Synopsis: A study of thermal spa baths for chronic low back pain revealed some benefits over warm tap water baths. However, multiple hypothesis testing takes away from the confidence clinicians can have regarding the true effect. Nonetheless, all patients improved during and after the treatment period of daily baths.
Source: Kulisch A, et al. Effect of thermal water and adjunctive electrotherapy on chronic low back pain: A double-blind, randomized, follow-up study. J Rehabil Med 2009;41:73-79.
This randomized, double-blind, controlled trial compared the effectiveness of thermal mineral water with tap water in the treatment of chronic low back pain. The study was conducted at the spa of Celldömölk in Hungary, which opened in 2005. Seventy-one patients with chronic lumbar pain (more than 12 weeks duration) received 20-minute treatment sessions daily for three weeks. Participants were randomized into two groups, which used baths filled either with medicinal spa water or with tap water, both at 34° C. Everyone also received standard electrotherapy at the same time, a treatment popular in continental Europe and similar to electrical stimulation that uses long pulse durations. In this study, electrodes were applied to people's waists, but few other details of the therapy were given.
Outcome measures were four visual analogue scales (VAS) for various back symptoms, range of mobility using Schober's test and the Domján's R and L tests, lumbar spine function using the Oswestry index, and quality of life using the Short Form-36 health survey. The tests were administered at baseline, immediately after the three weeks of treatment, and 15 weeks later.
After the three-week treatment period, the thermal water group showed significant improvements in all parameters compared to baseline. These improvements were maintained after 15 weeks of follow-up. All parameters also improved significantly in the control group compared to baseline, but to a smaller extent than in the thermal water group. Between-group comparisons revealed some statistically significant differences. Immediately after the treatment period, the thermal water group had significantly better scores on one VAS score. After follow-up, significant between-group scores existed on one of the other VAS scores and the Schober's index. These results were analyzed according to the intention-to-treat principle. When only those patients who completed the study were analyzed, other study parameters showed significant differences favoring the thermal water bath. The authors' conclusion was: "In the group treated with thermal water, improvement occurred earlier, lasted longer, and was statistically significant."
This study was well-designed and clearly reported. The inclusion and exclusion criteria for patients were clearly described, with anyone receiving physiotherapy at the same time excluded to avoid confounding effects. The randomization was conducted using a random-number table. An extract of green walnut husks was added to the control baths to make them resemble the thermal water, and all the baths were located in the same hall to provide the same odor and environment.
However, the study's results are limited by a number of factors not discussed by the investigators. The authors noted that the improvements reported were smaller than expected, but they did not report what those expected values were or the results of a power calculation. The 71 participants may have been an insufficient number to detect true differences. In addition, 20 participants did not complete the study, giving a relatively high drop-out rate of 28%. The impact of this can be seen in the sub-group analyses carried out by the investigators. To their credit, their primary analysis was carried out according to the intention-to-treat principle and revealed significant differences in three outcome parameters. When the data for only those who completed the full protocol were analyzed, an additional six parameters were significantly different. While the investigators found this gave credence to the value of the thermal baths, such an analysis inflates the true effect of an intervention and is why the intention-to-treat analysis is used.1
An even bigger problem with this study is the large number of outcomes measured. Multiple hypothesis testing is one of the most common problems in clinical research.2 When two or more hypotheses are tested, the probability of finding a positive result by chance alone is increased. For example, if five tests are used, there is almost a one in four chance that a positive result will be found by random chance. This study used nine tests, each implemented on two occasions, making it highly probable that some positive findings would result. None of the tests was selected as the primary outcome. Testing several hypotheses may be legitimate, but in such cases additional statistical analyses should be used to correct for potential problems.3 These were not reported in this study.
This interesting and well-designed study demonstrates the benefit of warm baths for chronic low back pain. Both the control group and the thermal water group showed improvements during the three weeks of treatment that were sustained for 15 weeks. However, even this conclusion must be held tentatively as a no-intervention control group was not included in this study. The small number of subjects, high drop-out rate, and numerous hypotheses tested mean that little confidence can be placed in the investigators' conclusion that thermal baths are more beneficial than warm baths with tap water. Given that many more of the parameters tested showed no differences between the two groups, patients can be reassured that they probably do not need to travel far from home to get the benefits of a warm bath.
1. Busse JW, Heetveld MJ. Critical appraisal of the orthopaedic literature: Therapeutic and economic analysis. Injury 2006;37:312-320.
2. Goodacre S. Critical appraisal for emergency medicine 2: Statistics. Emerg Med J 2008;25:362-364.
3. Bono CM, Tornetta III P. Common errors in the design of orthopaedic studies. Injury 2006;37:355-360.