“We should have listened to Brian!” Or “The Python’s knew all along”.

For decades we have known that treatment outcomes for people with low back pain have been suboptimal. The research community has been trying to determine subgroups, based upon key individual characteristics, amongst those with low back pain. The aim behind such subgroup-based approaches is that clinicians can identify which subgroups patients are in and then provide treatments that are individually matched to the patient based on their subgroup. Together, this process is hypothesised to offer better clinical outcomes.

In 1975 Graham Chapman led his Monty Python knights on the search for the holy grail. At approximately the same time-point researchers began exploring the importance of low back pain sufferers’ personality to clinical outcome.[1] Since then subgrouping has been the holy grail for many back pain researchers. Most subgrouping systems are based upon one dimension of a person’s presentation – e.g. – movement-related subgroups, subgroups with differing psychological profiles, etc. Unfortunately, when treatments are matched to subgroups based upon one dimension, it appears that outcomes fare no better.[2-4]

As our understanding of low back pain deepened, it became clear that it is a highly complex disorder where central and peripheral nociceptive processes are influenced by factors from multiple dimensions. The holy grail then shifted to trying to derive multidimensional subgroups.

In parallel to our work, a team in Denmark statistically-derived subgroups from multidimensional data from over 900 people with low back pain.[5] While the previous sample derived subgroups, in our sample that evaluated 294 people and 66 variables – e.g., movement, pain sensitivity, psychological, social, pain characteristics, demographics, health, lifestyle – we were unable to statistically derive subgroups from this data. We interpreted our finding as possibly being reflective of the high heterogeneity and complexity of our data.

So what next? We statistically derived subgroups based on three clinically modifiable and important single dimensions that have support from the literature. There were: 1) pain sensitivity[6]; 2) psychological questionnaire scores[7]; 3) pain responses following repeated spinal bending[8]. In addition, we explored the broader multidimensional profiles of the unidimensional subgroups. For instance, in the study examining psychologically-derived subgroups[7], those with high psychological questionnaire scores also had the highest proportion of participants with pain summation following repeated spinal bending and significantly greater sensitivity to pressure in the lumbar region. So, we got to thinking, could clinicians assume that those with high pain sensitivity will also have high psychological distress and greater pain summation following repeated spinal bending?

To explore this, we looked at patterns of presentation at the level of the individual across the three unidimensional subgrouping studies[9]. Because we happened to have three subgroups in each of three studies, given all the possible unique combinations, this resulted in 27 possible unique patterns of presentation.

We found 26! And looking at one extreme presentation – those with high pain sensitivity, high psychological questionnaire scores, and summation of pain following both forward and backward repeated movements – only 2.7% of the sample met this presentation.

This study suggests clinicians cannot assume that a presentation within one dimension (e.g., high psychological questionnaire scores) will be associated with a presentation in another dimension (e.g., high pain sensitivity). Such results suggest that unidimensional subgrouping is unlikely to capture the complexity of low back pain, possibly helping explain the modest results of treatments based on unidimensional subgroups. Indeed, clinicians may need a flexible clinical framework to characterize individual’s multidimensional presentations and guide management (e.g.[10])

In 1979 Graham Chapman acted out the Life of Brian. Opening the balcony doors and addressing his followers he shouted, “You don’t need to follow anybody. You’ve got to think for yourselves. You’re all individuals.” It seems the Python’s were ahead of the back pain world – giving up the holy grail and realising the high levels of individuality in people’s presentations.

About Martin Rabey

Martin is a Specialist Musculoskeletal Physiotherapist (As awarded by the Australian College of Physiotherapists, 2009). He completed his PhD at Curtin University in Perth, examining the complex interactions between factors which make low back pain persist. He recently moved back to Guernsey in the Channel Islands, where he was born. There he is a Director of THRIVE, a company with a three-pronged approach: offering education to healthcare practitioners worldwide on the integration of pain science with clinical practice and undertaking international collaborative research into persistent pain disorders, all the while delivering expert physiotherapy to the islanders. In his spare time he is a coasteering guide.

References

[1] Freeman C., Calsyn D., Louks J. (1976) The use of the Minnesota Multiphasic Personality Inventory with low back pain patients. Journal of Clinical Psychology 32: 294-298.

[2] Apeldoorn A, Ostelo R, van Helvoirt H, Fritz J, Knol D, van Tulder M, De Vet H. A randomized controlled trial on the effectiveness of a classification-based system for sub-acute and chronic low back pain. Spine 2012;37:1347–56.

[3] Bergbom S, Flink I, Boersma K, Linton S. Early psychologically informed interventions for workers at risk for pain-related disability: does matching treatment to profile improve outcome? J Occup Rehabil 2014;24:446–57.

[4] Henry S, Van Dillen L, Ouellette-Morton R, Hitt J, Lomond K, DeSarno M, Bunn J. Outcomes are not different for patient matched versus nonmatched treatment in subjects with chronic recurrent low back pain: a randomized clinical trial. Spine J

[5] Nielsen A., Kent P., Hestbaek L., Vach W., Kongsted A. (2017) Identifying patient subgroups using latent class analysis: should we be using a single-stage or two-stage approach? A methodological study using a cohort of patients with low back pain. BM Musculoskeletal Disorders 18: 57.

[6] Rabey M, Slater H, O’Sullivan P, Beales D, Smith A. (2015) Somatosensory nociceptive characteristics differentiate subgroups in people with chronic low back pain: a cluster analysis. Pain 156: 1874–1884.

[7] Rabey M, Smith A, Beales D, Slater H, O’Sullivan P. (2016) Differing psychologically-derived clusters in people with chronic low back pain are associated with different multidimensional profiles. Clinical Journal of Pain 32: 1015–27.

[8] Rabey M, Smith A, Beales D, Slater H, O’Sullivan P. (2017) Pain provocation following sagittal plane repeated movements in people with chronic low back pain: associations with pain sensitivity and psychological profiles. Scandinavian Journal of Pain 16: 22–8.

[9] Rabey M, Smith A, Kent P, Beales D, Slater H, O’Sullivan P (2019) Chronic low back pain is highly individualised: patterns of classification across three unidimensional subgrouping analyses. Scandinavian Journal of Pain (In Press).

[10] Mitchell T, Beales D, Slater H, O’Sullivan P (2019) Musculoskeletal Clinical Translation Framework. https://www.musculoskeletalframework.net/ (Accessed 1.7.19)