No paper was found concerning patients at risk for being progressive, followed to skeletal maturity under physiotherapy treatment alone. Claims made to regard physiotherapy as an evidence based method of treatment, are therefore, scientifically unjustified [[10, 12] Romano et al. Cochrane Review in press].
The only evidence on Level II is found in the immature sample (n = 94) from the prospective controlled study from our group [23]. However, this group of patients was not followed up to skeletal maturity.
Another paper comparing two unproven groups of physiotherapy (general exercises vs. SEAS [Scientific Exercise Approach to Scoliosis], described in [12] against each other) does not seem to provide any evidence as this study design does not make sense [16] because the differences found between the two groups cannot necessarily be regarded as leading to the conclusion that one of the therapies might be of any benefit to the patients treated:
When one method is not effective and has no better results than observation only, the other method could also lead to deterioration and therefore be statistically different (Figure 4). This is the cause why only controlled studies with an untreated control group can be regarded as a valid source of scientific information.
Nevertheless, the authors claim in their conclusion: This data confirms the effectiveness of exercises in patients with scoliosis who are at high risk of progression. Compared with non-adapted exercises, a specific and personalized treatment (SEAS) appears to be more effective [16].
This study included seventy-four consecutive outpatients with adolescent idiopathic scoliosis, mean 15 degrees (standard deviation 6) Cobb angle, 12.4 (standard deviation 2.2) years old, at risk of bracing who had not been treated previously [16].
Italian girls with an average age of 12.4 years surely can be estimated as being postmenarchial with a Risser 1 at least. When calculating the average patient from this sample using the Lonstein and Carlson formula [4, 5] with a curvature of 15 degrees one can estimate a risk factor of < 1 and therefore this patient sample is not at risk, but is a benign sample of not needing any treatment because the risk for progression does not exceed (actually is < 15%) 40%, which according to the guidelines would be the indication for physical treatment. But even if the Risser sign is assumed as unlikely to be 0, there was no indication for treatment as the progression factor would still be 1.2, only, thus still below 40%. So also this calculation would lead to the conclusion of clear unnecessary overtreatment and in no way the patient sample is at risk of needing a brace as stated in the paper.
This study is only one example of documented malpractice and unfortunately literature is full of samples not needing any treatment, but claims have been made from these studies that physiotherapy would be of benefit.
The authors have also published a retrospective study [24] including a „worst case analysis "with a patient sample (n = 112) of 13.2 years and a Cobb angle of 23.4 degrees. Also this sample lacks any indication for treatment (23.4-3 × Risser 2 (estimated benign, because many Mediterranean girls have Risser 3 at the age of > 13 years, as have German girls too) /13.2 = 1.32 , which makes less than 40% chance for progression with a clear indication for observation, only) As these girls surely are in the descendent phase, the prognosis is getting better every day and as a curve of < 30° is very unlikely to progress after cessation of growth, there is surely no urgent indication for treatment and no real risk for the need of brace treatment).
The Chinese RCT [25] has a patient sample (n = 80) with an average age of 15 years at the start of the follow-up period and a follow-up time of 6 months at an average. 15 year old girls (girls are the main population in samples with AIS) usually do not have significant residual growth left and do not necessarily need any treatment. So this study, even with the most important study design (RCT) cannot contribute to the search for evidence for PT in scoliosis (Table 1). Additionally there is no evidence that the curvature does not return to the initial value after the period of physiotherapy.
The problem of treating mature patients and claiming beneficial outcomes is also evident in bracing [26] (Figure 5).
As can be seen in Table 1, only 7 out of 19 samples of patients published [23–25, 27–41] had a risk of progression exceeding 40% and by way of this had an indication for treatment (38%). One study had a short-term pre-post design and should be excluded [27]. Three other papers were with patient samples (n = 80 [25], n = 69 [31], n = 53 [32]) that were (nearly) outgrown and would not need any treatment [25, 31, 32].
The studies by Weiss 2003, Mollon and Rodot 1986 and Ducongé 2002 [23, 29, 34] with respect to the materials included were not homogenous, had a wide range of materials and included also many prepubertal patients not yet at risk. Therefore finally only 4 out of the 19 samples [23, 35, 37, 41] can strictly be regarded as having had an indication for physical therapy. However, all these 4 samples were postmenarchial when the observation started at the descendent part of the pubertal growth spurt.
No controlled paper with adult patients at risk for progression (curvatures exceeding 35°) [2] has been found.
According to the findings from this review, studies on physiotherapy in idiopathic scoliosis patients have the following shortcomings:
Prospective study designs should not be overestimated when the material within the study can be inappropriate [42]. In studies on scoliosis this is the case when mature samples without any treatment indications are studied using prospective controlled or even randomized designs (e.g. [25]). Maybe there is a benefit also for this population from applying PT, but only a patient sample at risk for being progressive in the well renown range of proper treatment indications can be accepted and can contribute to evidence in this field. This may also be the problem within some Cochrane reviews [43].
It is also important to see the current evidence for physiotherapy during growth within the context of the other module of conservative treatment such as bracing. The only paper presenting at least some evidence for physiotherapy is the prospective controlled paper from my previous working group [23], however, this patient sample was not followed up to the end of growth (maturity). Within this study we find a subsample of patients at higher risk for being progressive. The controls from this study were non progressive in 30% without any treatment which compares well to the controls in the SRS brace study [7]. In immature patients intensive inpatient physiotherapy can halt progression in 50%, however the Boston brace without PT will be effective in 70% [7]. The Chêneau brace of the 1999 standard is effective in 80% [9] while today effectiveness has increased to > 90% [44]. Therefore, bracing seems to be the most important approach in the conservative management of patients with scoliosis during growth (Figure 6).
Recent papers present samples not followed up by x-ray investigations and therefore could not be compared to other studies from this review [20–22]. However, in at least two of these studies [21, 22] mature patient samples were studied (and in part treated as in-patients for many weeks), who were not at risk for deterioration.
Thus overtreatment seems to be an important issue in the studies on conservative management of scoliosis.
There are a few studies on conservative treatment of adult scoliosis patients published recently [17–19], but no studies with an untreated control group. Although the limitations of these studies were discussed, the authors draw conclusions even though their studies have major shortcomings. In one of the studies the authors [17] state themselves: An important caveat of this study was that the treatment was not randomized and therefore the treatment group might have deteriorated if not for the treatment they received. Bridwell et al. [18] had drop-out rates of more than 50% in the non-operative group, so no conclusions are justified from this paper, because a ‚worst case' analysis would possibly lead to the opposite conclusions.
A similar paper, published in 1995 [45] was also accepted for publication in the American edition of the Journal of Bone and Joint Surgery, although the conservative sample had a return rate of 50% only. Scientifically these papers do not merit being published as their material is poor and the conclusions drawn, invalid.
Study design may be prospective controlled, however as the problems (complications) of spinal surgery arise after many years, mostly with a lifetime risk of 40-50% [2, 46], a follow-up of two years seems very questionable and therefore these papers comparing surgery to conservative treatment of scoliosis seem no valid source of information.
An agreement of the scientific community on common inclusion criteria for future studies on PT seems necessary. We suggest the following: (1) girls only, (2) age 10 to 13 at the first signs of maturation (Tanner II), (3) Risser 0-2, (4) risk for progression 40-60% according to Lonstein and Carlson.
It would be even better to only include patients from the ascendant phase of the pubertal growth spurt: (1) girls only, (2) age 10 to 12 at the first signs of maturation (Tanner II), (3) Risser 0, (4) risk for progression 40-60% according to Lonstein and Carlson.
However, the problem would be that the number of the patients included would be very limited as has been shown in a prospective controlled paper on bracing using this kind of inclusion criteria [9]. Usually scoliosis in adolescent girls is detected after the onset of menarche and therefore the suggestion also including patients with Risser 1 or 2 seems to be more reasonable.
The postural correction plays a major role in physiotherapy like in bracing [7–9, 47] treatment and can be achieved for instance by side shift exercises or the recent developments of the Schroth method [31, 32, 48, 49]. The methods for exercising (Yoga, Pilates, SEAS, DOBO MED) presented in the review by Fusco et al. [12] are not sufficiently evaluated and should be questioned.
There is still no evidence that physiotherapy exercises can decrease the progression of scoliosis in immature samples with idiopathic scoliosis (with significant Cobb angles > 15 degrees) and thus the correction by braces is emphasized. However, physiotherapy exercises should be regarded as a complement to bracing concerning postural control during activities of daily living (ADL) [49]. Postural experience and postural correction are important to stimulate a good posture in grown-up individuals. A specific method of teaching the patient to achieve an optimal postural control was introduced by Schroth and optimized recently [49]. The positive effect of physical exercise on peak bone mass and on balance performance/coordination in growing children/adolescents should of course not be underestimated [50, 51].
PT may have a beneficial effect on the patient with idiopathic scoliosis as this has been demonstrated in many pre-/post cohort studies [12], however during the most vulnerable period of the pubertal growth spurt PT should never be regarded as the only meaningful mode of treatment [52].
