Brief psychosocial education reduced the incidence of low back pain?

Can a brief psychosocial education reduce the incidence of low back pain?
By Terra Rosa

A paper published in BMC Medicine in 2011 “Brief psychosocial education, not core stabilization, reduced incidence of low back pain: results from the Prevention of Low Back Pain in the Military (POLM) cluster randomized trial” has made a bit of headline and controversy among manual therapists. This paper is an open access manuscript, and can be read here

The title of the paper suggests that core stabilization doesn’t work, but a brief one-off ‘psychosocial’ education session can lower the incidence of low back pain (LBP). People then start to jump into a big conclusion, and make claims that specialized core training is a waste of time, is no better than traditional exercise, but one session of 45 mins of ‘psychosocial’ education can reduce the incidence of low back pain by up to 5%!

Is that really that powerful? You sit in a seminar room and got lectured for 45 mins about low back pain, its anatomy, its evidence-based treatment and how to avoid it. Then somehow this ‘psychosocial’ education ‘stayed’ in your brain and even after two years it can reduce the incidence of low back pain.

Seems too good to be true, but from my statistical re-analysis of the data, the results are really not that great. (I have the statistical background to do this analysis, see below for more details or  read this document LBP data analysis)

My analysis suggested that the differences in LBP incidence between all 4 treatments are small (traditional lumbar exercise (TEP), traditional lumbar exercise with psychosocial education (TEP+PSEP), core stabilization exercise (CSE), or core stabilization with psychosocial education),  and not statistically different.

Details of Statistical analysis

From the paper, the clinical method was described as follows:

“The Prevention of Low Back Pain in the Military study was a cluster randomized clinical study with four intervention arms and a two-year follow-up. Participants were recruited from a military training setting from 2007 to 2008. Soldiers in 20 consecutive companies were considered for eligibility (n = 7,616). Of those, 1,741 were ineligible and 1,550 were eligible but refused participation.

For the 4,325 Soldiers enrolled with no previous history of LBP average age was 22.0 years (SD = 4.2) and there were 3,082 males (71.3%).

Companies were randomly assigned to receive

• traditional lumbar exercise (TEP), 
• traditional lumbar exercise with psychosocial education (TEP+PSEP),
• core stabilization exercise (CSE), or
• core stabilization with psychosocial education,

The psychosocial education session occurred during one session and the exercise programs were done daily for 5 minutes over 12 weeks.The primary outcome for this trial was incidence of low back pain resulting in the seeking of health care.

And here are the Results:

“There were no adverse events reported.

Evaluable patient analysis (4,147/4,325 provided data) indicated no differences in low back incidence resulting in the seeking of health care between those receiving the traditional exercise and core stabilization exercise programs.

However, brief psychosocial education prevented low back pain episodes regardless of the assigned exercise approach, resulting in a 3.3% (95% CI: 1.1 to 5.5%) decrease over two years (numbers needed to treat (NNT) = 30.3, 95% CI = 18.2 to 90.9).”

Now, here is when statistical analysis can be interpreted differently:

“Results from the generalized linear mixed model indicated that Soldiers in the combined exercise and psychosocial education groups (CSEP + PSEP and TEP + PSEP) were similar,
but experienced an average of 0.49 fewer months with incidence of LBP (95% CI: 0.003 to 0.983, P = 0.048) in comparison to those not receiving PSEP. “

I analysed the data from Table 3 of George et al. (2011). Since I do not have all other data, I did a simple linear mixed model:

Percent LBP incidence = Treatment (Fixed effect) + company (random effect)

The model said that the percent of LBP incidence is a function of the treatment (which we called fixed effect) and company (called random effect, because the company or group of people can behave differently). This document (LBP data analysis) shows the results of my statistical analysis.

The results of my statistical analysis: While the model says that TEP + PSEP will decrease the mean of incidence by 2% decrease , they are not statistically different from other treatments. All the parameter estimates from the linear model indicated that the coefficients are not statistically significant. Now we can do another statistical test to see if there are differences between treatments.

The first is the usual Student’s t-test at the level of 0.05 (95 percent confidence interval, or 95% of the time it does not happen by random chance).

Although TEP+PSEP has a lower mean (14%), it is not statistically different from CSEP+PSEP. In addition, core stabilisation (CSEP), traditional exercise program (TEP), and (CSEP + PSEP) are not statistically different.

If we used another more rigorous statistical Tukey HSD test, which compares all the treatments, the result shows that the means for all treatments are similar, or not significantly different!

Since I do not have the raw data, I can only analyse what was given in the paper, the percentage of LBP incidence (after 2 years of trial). My analysis suggests that the differences in LBP incidence between all 4 treatments are small, and are not statistically different.

The authors then made another step in generalising the psychosocial education (PSEP), by collapsing the treatment into 2 groups! Having psychosocial education (yes or no) or core stabilization (yes or no). Really, there are 4 separate treatments in different groups of people, and you can’t simply group them for ‘more efficient communication of results’. This should make PSEP stands out, but even with this grouping, the difference between CSPE and PSEP is only 1.8% (CSEP mean incidence of LBP 16.7%), and psychosocial education (PSEP mean incidence 14.9%). Then the authors ‘adjust’ the baseline and suggested that PSEP was estimated as having 3% lower LBP incidence.

“The analyses of the four intervention groups suggested a pattern that allowed for more efficient communication of results by collapsing the intervention groups into those receiving any core stabilization (CSEP – yes or no) or any psychosocial education (PSEP – yes or no). There were no differences between the TEP + PSEP and CSEP + P
SEP groups, but chi-square test indicated that receiving the PSEP program with any exercise program was protective of LBP incidence (Chi-square = 5.56, P = 0.018; and 5.05, P = 0.025 when adjusted for intracluster correlation) in comparison to those not receiving PSEP. Furthermore, after adjusting for demographic and baseline levels of clinical variables, the protective pooled effect of any PSEP was estimated at 3.3% (95% CI: 1.1 to 5.5%) decreased LBP incidence over two years (P = 0.007). This effect corresponds to numbers needed to treat (NNT) of 30 (95% CI = 18.2 to 90.9).”

Readers interested in the science and the process of an article review, should also read the comments by reviewer Dr. Raymond Ostelo who expressed a skepticism on the effect of a 1 session of education and questioning the “collapsing” intervention group: http://www.biomedcentral.com/imedia/1391493222615742_comment.pdf

I am not sure what to make out of this, but my view is that the authors have stretched the results a bit far and made a bold conclusion that a brief psychological education can lower the incidence of LBP. Let’s be serious if you are briefed in a 45 mins session on the evidence-based treatment of LBP, will you still remember it after a month? The statistical difference is small and probably makes not much of a difference.

There are various limitations in this study, as it was carried out on a military population, and the results’ differences are not huge. Scientists nowadays like to made a bold controversial statement, sometimes stretching the results a bit further. I once read a suggestion that a bold statement …can add a little spark to your paper.

Also consider the conclusions other LBP studies conducted by the same group of authors (using the same military group). There are many other factors that influenced the results.
Effects of traditional sit-up training versus core stabilization exercises on short-term musculoskeletal injuries in US Army soldiers
http://www.ncbi.nlm.nih.gov/pubmed/20651013

“soldiers who completed the TEP and experienced a low back injury had more days of work restriction: 8.3 days (SD=14.5) for the TEP group and 4.2 days (SD=8.0) for the CSEP group.”
Predictors of Occurrence and Severity of First Time Low Back Pain Episodes: Findings from a Military Inception Cohort

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0030597#pone.0030597-Childs1

In this article, the authors said:
“Education level and physical fitness were consistent predictors of pain intensity, while gender, smoking status, and previous injury status were predictors of disability. Gender, smoking status, physical health scores, and beliefs of back pain were consistent predictors of psychological distress. “

OK, so is psychosocial education not important? I didn’t say so, but the authors’ suggested 3 references that deal with this issue. And what can be learnt is that additional psychosocial educational materials can be supplemented to manual therapy.

George SZ, Teyhen DS, Wu SS, Wright AC, Dugan JL, Yang G, Robinson ME, Childs JD:Psychosocial education improves low back pain beliefs: results from a cluster randomized clinical trial NCT00373009) in a primary prevention setting.
Eur Spine J 2009, 18:1050-1058. PubMed Abstract | PubMed Central Full Text

In this article, the same authors recruited companies of soldiers (n = 3,792), and cluster randomized to receive a PSEP or no education (control group, CG). A back beliefs questionnaire (BBQ) was given, which assesses inevitable consequences of and ability to cope with LBP. The BBQ was administered before randomization and 12 weeks later. The authors concluded that soldiers that received the PSEP had an improvement in their beliefs related to the inevitable consequences of and ability to cope with LBP.

George SZ, Fritz JM, Bialosky JE, Donald DA: The effect of a fear-avoidance-based physical therapy intervention for patients with acute low back pain: results of a randomized clinical trial.
Spine 2003, 28:2551-2560. PubMed Abstract

In this article, the authors concluded that “Patients with elevated fear-avoidance beliefs appeared to have less disability from fear-avoidance-based physical therapy when compared to those receiving standard care physical therapy. Patients with lower fear-avoidance beliefs appeared to have more disability from fear-avoidance-based physical therapy, when compared to those receiving standard care physical therapy. In addition, physical therapy supplemented with fear-avoidance-based principles contributed to a positive shift in fear-avoidance beliefs.”

Coudeyre E, Tubach F, Rannou F, Baron G, Coriat F, Brin S, Revel M, Poiraudeau S: Effect of a simple information booklet on pain persistence after an acute episode of low back pain: a non-randomized trial in a primary care setting.
PLoS ONE 2007, 2:e706. PubMed Abstract | PubMed Central Full Text

In this article, the setting is in a primary care practice in France. Participants: 2752 patients with acute LBP. Intervention: An advice book on LBP (the “back book”). Main outcome measures: The main outcome measure was persistence of LBP three months after baseline evaluation. The conclusions: “The level of improvement of an information booklet is modest, but the cost and complexity of the intervention is minimal. Therefore, the implications and generalizability of this intervention are substantial.”