This is the first research study that has looked at the role of a complete treatment team, as recently focused in the literature , in the management of braced patients. Following an exploratory analysis we performed a case–control retrospective study (first step of prognosis studies) that showed astonishing odds ratios for the studied risk factor (presence or absence of a complete team according to the SOSORT criteria :) 5.5 times increase in non-compliance, 15.7 of QoL disturbances and 16.8 of pain in patients not surrounded and helped by adequate teams.
The immediate answer to bracing clearly shows how this treatment is stressful, with no differences between all the considered groups. Nevertheless, TEAM, compliant and pain-free patients recovered in less time and much better than NOT, non-compliant and painful ones. This means that something helped the first groups more than the second ones. According to the results of this study, the specific help could be neither the physician (prescriptions, treatment management, checking of the brace) nor the orthotist (construction and management of the brace), since they were the same (and behaved in the same way), but it was either the team as a whole or the APs. Consequently, the QoL was dramatically increased, pain decreased and the usage of brace increased. Ultimately we could also expect better results to bracing (due to compliance [13, 14]) even if the actual QoL and pain are results “per se.”
As already stated, due to the design the physician and orthotist can be excluded as determining factors of the differences found in the two studied groups. However, when we consider the two different settings as the possible explanation of these results, we should look not only at the organization and collaboration of the team, but also the setting of physiotherapeutic approach (private versus HNS) that could drive certain social differences. Nevertheless, if it is hypothesized that only the wealthiest treat their children in private institutes, it is not necessarily true that the richest are the most compliant with mildly invasive procedures like bracing (it is even possible the contrary); on the other hand, to face at best a stressful event like bracing requires adaptability and elasticity together with external support [23–26]. Even if we found no difference at the baseline for the groups considered, we cannot exclude that people requiring help in a private setting are at the start different from those who seek help from the NHS, and this point should be considered in future prospective studies.
We directed our attention with this study mainly to the team surrounding the patient and particularly on APs. In fact, beyond what they do (i.e., the types of exercises), how they behave and what they say are extremely important. While physicians have the authority of leadership, prescription and indications, while orthotists have the intensity of helping patients in the first impact with the brace, APs play a major role as well, due to their continuous presence at the patient’s side. APs neither prescribe nor build braces, but consequently they appear to the patient as a third, expert judge. Moreover, they have time, due to their continuous weekly work, to explain, sustain, drive and help the patients and their families in a way they can be much more important for the team than the others do. However, they can also be much more destructive: Words like “I would never wear that brace!” or “To me, braces destroy muscles and should not be used” or similar can completely undermine the work of the other members of the team. This is one of the main complaints of the physician and orthotist who participated in this study when considering the APs involved in NOT. This can drive the patients to increased problems and difficulties, as shown in this study, even if they do not perceive this negative impact. In fact, the patients in TEAM and NOT did not perceive their treatment teams differently.
In this study we did not have a group of patients treated in a team in which no AP was involved, and theoretically that should be the real and “pure” control group. Nevertheless, in this situation we would also add the difference due to exercises to that of the team composition. Because we know that exercises do have a favorable effect on scoliosis patients [27, 28], and specifically on those braced , the actual study is presumably the best way to check for the team role only. Nevertheless, even if we believe this is not true, we cannot ignore that the differences found could be due to a negative effect of the APs in NOT more than to a positive effect of those involved in TEAM.
When we started this study, we mainly hypothesized the possibility that the absence of a good team could incur reduced compliance, while we were not really concerned with QoL issues. Thus we proposed the SRS-22 only for the purposes of checking. Only after the exploratory research did we add pain as a possible outcome. The final results on pain and QoL were very impressive and, in a way, much more important than those on compliance. In fact, as stated by SOSORT experts, QoL and disability are among the main aims of treatment, being more important than Cobb degrees . In that respect compliance should “only” drive better final results in Cobb degrees, which should ultimately correlate with future QoL. On the contrary, the “actual” QoL is always reduced by bracing [1, 2], and if the team is able to decrease this reduction it should be very welcome. The same is completely true in regard to pain: Patient management plays a major role in their pain perception, as it is already well known in the literature, mainly for adults, where low back pain is concerned [31, 32]. Consequently, the team role appears to be even greater in this study for QoL and pain issues than it is for compliance.
About pain, it is interesting to see that the pain scale in the SRS-22 was not statistically different between the two groups (Table 5), while the difference was noted in the other questionnaire we proposed (Table 4). We must note here that the SRS-22 explore the everyday pain experience mainly related to scoliosis, and not to bracing. Conversely, the questionnaire we developed asked what was the most important problem perceived by patients during brace wearing.
Interestingly, the response rates were different in the two groups. We were very careful in the administration of the questionnaire to guarantee anonymity, with the drawback that it was not possible to check who responded to the questionnaires. Moreover, it was not possible to specifically solicit the answers of non-responders. We consider the reduced response rate as another indication of the greater compliance and better team approach in TEAM than NOT: In fact, responders should be more numerous if patients are highly motivated. Moreover, when we consider that the best patients prefer to show how good they are, an “intent-to-evaluate” analysis, in which we would consider as “failure in compliance” those who did not answer, would only increase the differences found in favor of TEAM. The main limitations we must consider in this study include the following:
Design: A case–control retrospective study allows one to find correlations with possible risk factors but not to draw cause-effect relationships, for which future prospective cohort studies, including a logistic multiple regression analysis, should be planned. Consequently, here we have odds ratios and neither relative nor absolute risks. Nevertheless, in the absence of any study this design is the appropriate first step toward a better understanding.
Selection bias: As discussed above, it is possible that auto-selected patients according to the choice of being treated in a private institute (versus the NHS) are more compliant for other characteristics than the treatment team would be, and this requires other designs to be solved; moreover, in a retrospective study it is not possible to check for patients who abandoned treatment. In any case, we found no differences at the baseline between the groups.
Confounding bias: We controlled for this, but nevertheless we must consider that pain, QoL and compliance could be interrelated and one could drive the others. Only a multiple logistic regression analysis coming from a prospective study will in the future make it possible to deeply check this possibility.
Population: Reduced sample. This is due to the difficulty of finding a population treated in two separate teams while maintaining the same physician, orthotist, and exercises approach. This did not allow sub-analysis to be performed. However, and conversely, due to the reduced statistical power, reaching the statistical significance meant showing solid results.
Population: Various pathologies included. This was necessary in order to have the sufficient numbers needed to reach at least some conclusions. Since we found no difference between the two studied groups for the factor of “disease,” it should not count in the final results. Moreover, we were interested in finding the answers to the stressful event “bracing” in adolescents with spinal deformities instead of a specific disease. According to the clinical experience of clinicians working in TEAM, pain is really rare in braced patients, either for scoliosis or for kyphosis. In fact, the results of this study were surprising for them. As a consequence, results presumably really come from the different settings.
Response rate: Low in NOT. This could interfere with the results even if we explained, as above, the possibility that they could reduce the differences we found.
Compliance measurement: We did not use a compliance meter but instead used only a questionnaire. Nevertheless, we guaranteed complete anonymity to patients, and presumably that allowed us to obtain reliable answers, provided the existing and known “gap” between real and referred use of braces .
QoL at baseline. Unfortunately we did not have a QoL measurement at baseline, and consequently we cannot study the differences but only the actual values of the SRS-22 questionnaire.
Conversely, the strengths of this paper include the fact that this is the first study to have looked at the treatment team as a possible factor driving the results: This is a topic that is relatively “hot” in the literature, since it has been underlined only recently . Moreover, as already stated above, the reduced sample considered, with its low statistical power, allows us to state that the differences we found are very strong.
In this respect, it could also be argued that the study population is too small to be representative. It must be considered that we have been highly selective in choosing the population so to respect the inclusion criteria required by the study. In fact, it is very difficult in the everyday clinical life to have a situation where it is possible to explore different teams with the same physician and orthotist involved. Usually physicians work in a single well defined setting, and not in two separate ones like in this specific case. In fact, this study was performed because it was in the interest of the physician to understand if the differences he was seeing between the two team settings were real or not: his aim was to increase the quality of his work in both settings according to the final results. So, when looking at the small population, we can on one side consider the work as preliminary to future studies, on the other very well focused on a specific, quite rare, clinical situation. Another final consideration relates the composition of the population. The overall population before the responses reflects some skew (even if not statistically significant) in the NOT group towards males (36%) and kyphosis (32%) vs (23% and 15%) for TEAM. The different response rates in the two groups (92% in TEAM vs 48% in NOT) could have intensified this skew. Since we know that girls show a higher compliance level to bracing in comparison with boys , and that patients with thoracic hyperkyphosis are significantly more symptomatic in all SRS-22 domains , it would have been important to know the final gender and disease allocation by group. Unfortunately it was not possible, due to the blind compilation of the questionnaires. Even if we did not find any difference (a part from function) in the SRS questionnaire between the two groups (to be expected if kyphosis males subjects were prevalent), future studies should address these points carefully.