In defense of adolescents: They really do use braces for the hours prescribed, if good help is provided. Results from a prospective everyday clinic cohort using thermobrace

This is a prospective cohort observational uncontrolled study focusing on everyday clinical use of a heat sensor (TB) to check compliance in patients braced for spinal deformities.

At brace prescription, if the physicians deemed it to be appropriate during the consultation, the patients’ families have been asked whether they wanted to check the compliance of their children to treatment through TB for clinical purposes: If this was the case, they had to buy TB, thus confirming their adherence to the project. TB was then applied by the orthotist during brace construction and the data was read by the treating physician at the next clinical evaluation. All parents of participating patients signed a consent form for clinical research. Because it was our institute’s first experience with the clinical introduction of the TB, some caution was applied at the beginning by the prescribing physicians, primarily for the purpose of avoiding any interference with a good physician-patient-family relationship. [27] Consequently, at the start it was proposed as complimentary (i.e. the patients and family had to decide if they want to buy it), and mainly to patients already in treatment who were showing problems in the use of the brace. Subsequently, it was also proposed as complimentary to the patients at their first evaluation at our institute, but not to all of them. At that time we did not have a real rule for inclusion/exclusion of the patients: exclusion were done only basing on factors like difficulties of patient/family/physician relationship of any kind; delays that precluded to have time to discuss also the choice of using TB or not; families who were arbitrarily thought that could have had economic problems; and so on. Only recently, following the results of this study and the clinical perception of its utility, it has become a standard clinical procedure in our institute (i.e., prescribed to all patients). Consequently, the prescription of the TB gradually increased with time, but with caution until the first TB results were received, in order to have time to check them as well as the possible consequent difficulties in managing patients. After this test period we began to understand that the regular use of this device should enrich our activity: that’s why we start the standard prescription of Thermobrace.

We report here on all the first patients who consecutively accepted the voluntary use of the TB for the clinical evaluation of brace usage. Sixty-eight adolescent idiopathic scoliosis (93%) or hyperkyphosis (7%) patients with brace prescriptions were recruited during our everyday clinical activity. There were 53 females, 15 males; the mean age was 14 years 6 months ± 2 years 3 months; scoliosis patients had maximum curvature of 41.4 ± 10.5 Cobb degrees at the start of treatment. In Table 1 the characteristics of these sub-groups are reported. All the study has been carried out with respect to the first group, while the second was used only to check whether referred compliance changed after starting the use of the TB.

Since this is an observational study of clinical everyday behaviors, it did not require approval of an ethics committee. All patients included in this study signed an informed consent to clinical data management for research purposes.

To measure the actual brace wear, we used a commercially available heat-sensor device called the “iButton™ DS1922L-F5#” (http://​www.​maxim-ic.​com/​datasheet/​index.​mvp/​id/​4088/​t/​al) (Maxim Integrated Products, Inc.; 120 San Gabriel Drive, Sunnyvale, CA 94086), which we called for this specific use “Thermobrace” (Figure 1). It is a small temperature data logger to be installed within each orthosis (Figure 2) constituted by a heat sensor, a battery and a memory. The sensor is meant to measure the brace temperature, and it is not placed in contact with the patient but within a pressure pad in the brace. Recently a reliability study has been published on this specific instrument [28].

Before starting the clinical application we performed a preliminary validation trial involving 5 female patients: They all had adolescent idiopathic scoliosis (range 35-53° Cobb), were recruited on a voluntary basis, and had been in brace treatment for more than 2 years. Two of them were already known for having bad referred compliance, while 3 had very good referred compliance. Since before this study we were not using any monitoring device, the only type of compliance we already knew of these patients was what they referred when inquired in front of their parents. Referred compliance has been shown to underestimate the real brace wearing [19].

We required all these patients to register on a data sheet each time they put on the brace or took it off, and we compared this data with that registered by the TB. We considered their data sheets as the golden standard reference to validate TB measurements. In fact, the data sheets were completely different from referred compliance: pupils had to fill in them during the day reporting exactly not the time of brace wearing, but the hour and minutes of the day so that calculation could be made by us afterwards. Moreover, patients did know that what we were searching for was the validity of the instrument and not their own compliance. Parents were asked not to check this data sheet, and we guaranteed we would have not used them to check their compliance (what we did, in fact). Finally, they had to perform this validation for one month, while the period of observation for their treatment was six months (so detaching compliance to therapy from the observation period of the pilot study).

The preliminary trial lasted 27.5 days (full memory): The TB showed good precision, with a difference between real and registered data ranging from 0 to 7.7% (Table 2). The two patients with the worst results were the non-compliant and had frequent brace on/off cycles. However, when the brace was worn regularly the reliability of the TB was very good (range 0–0.4% in compliant).

In the choice of the final time-span between each single measurement, we had to make a compromise. In fact, ideally the best solution would have been to measure temperature every minute, but there were memory limitation for an everyday clinical usage where patients are usually seen every 4 to 6 months (with some of them, usually low compliant, not coming back before 8–10 months). In fact, the more memory we used, the higher the costs of the device, and to let patients freely buy it we had to maintain this costs below 80–100 Euros, a threshold that was established after a preliminary informal inquiry with a number of patients. As a consequence, based on the recorded data (every 20 minutes) we checked the minimum time-span in order to discern the real use of the brace without losing precision between two clinical evaluations. After this trial we concluded that the TB was reliable and it was possible to sample the temperature every 60 minutes, thus optimizing the memory with good-quality data. It was also decided to develop a reliability test for clinical purposes, as explained below.

A specific software program with which to elaborate the data has been developed and is now freely available on-line (http://​www.​scoliosismanager​.​org/​thermobrace). To determine whether a sample corresponds to the brace being worn/not worn, it uses the following algorithm:

This method is sometimes not reliable with patients wearing braces 23 h/d. If the sampling period is too short (less than 3 months) or if the patient wears the brace all day with the exception of 20–30 minutes a day (which some of them do), the lower samples are not sufficient to reliably define the thresholds. This is why we defined the reliability index, to make the physician aware of potential problems. The reliability index evaluates many parameters, like the number of transitions brace on/brace off; degrees of interval between high threshold (brace certainly worn) and low threshold (brace certainly not worn); and number of samples. The index defines the reliability of the calculation of the threshold between brace worn/not worn (e.g., a patient who puts on or removes his/her brace many times during the day reduces the reliability).

Moreover, we defined another experimental calculation method: after collecting data of more than 50 patients, in different seasons (winter, summer), they were evaluated to find the mean thresholds of brace worn/not worn in their samples. We identified the thresholds as 28° (brace worn) and 25° (brace not worn). This second method simply assigned the samples with respect to those values.

In this specific study the data reliability was 84.0% (IC95 46.1-100) for the first of the two processing methods, versus 82.2% (IC95 39.4-100) for the second, thus indicating a statistically significant difference. Only 10 patients had better reliability results with the second method. In this study we consider only the most reliable data-processing method obtained in each single case (reliability: 84.3%; IC 95 46.1-100).

This specific software has been integrated with our clinical system (http://​www.​scoliosismanager​.​org). Today it allows clinicians to systematically access the following data:

Reliability of data (Figure 3a)

Environmental clime and temperature didn’t seem to affect data recorded by the heat sensor. This is a very challenging question and, due to reviewers suggestions, we analyzed our data in the attempt to investigate this possibility. So in a wider group of 312 patients (obtained in February 2012, after the original study was performed) we collected the moving mean obtained by the TB, for the high and low threshold in each months of the year. We looked for the hottest period of the year and we compared the average temperature per hour of the day with the mean brace wearing time per hour of the day in the same month. Moreover, we compared the mean moving average obtained from the TB in a city of the North of Italy with one in the South, as well as the hottest temperatures in the same period in these two towns: the trend was similar. The Figure 4 A and B shows the mean temperature per months of the year and the comparison of temperatures in Milan (North) and in Messina (South).

In the hottest period of the year (August) the wearing threshold has been passed for 208 hours, the worst possible effect is that a patient with a prescription of 23 hours per day, after a monitoring period of six months, may show a wearing time of 23 hours per day, measured by Thermobrace, while the real wearing time being of 21.8 hours per day. This is the worst possible case and the overestimation is quite little.

An additional indirect analysis can be done by checking the changes in wearing rates between the months of August and the nearest months (July and September): on average in August (the hottest month of the year), the brace has been worn 28 minutes more than in July (this could be an overestimation effect of the external temperature), but also 1 hour and 20 minutes fewer than in September (and this is in contradiction, since in Italy in September the temperature is considerably reduced when compared to August). Moreover, we have to consider that during summer people goes on holiday and have journeys, these differences are not significant and cannot be justified only by the hot weather.

Finally, with respect to the external weather, braces are used mainly at home, and in Italy most of the houses have air conditioning, so greatly limiting this possible, theoretic overestimation effect.

We considered two different types of compliance:

To calculate compliance, we considered the total amount of hours of brace-wearing from the brace check (TB mounting) to the first clinical follow-up (data discharge).

All patients have been treated with braces following the SPoRT principles [3, 4, 29‐31]: Sforzesco, Sibilla or Lapadula. The acronym “SPoRT” means “Symmetric, Patient-oriented, Rigid, Three-dimensional, active” and has been defined because these braces are designed to be as much as possible bodily shaped and adherent (Figure 2) so as to be easily masked underneath clothing; moreover, these braces allow complete movement with the limbs so as to permit sports activities, which are frankly encouraged [32]. All the patients also performed exercises according to the SEAS principles [4, 33‐35].

The setting in which treatment has been proposed fulfilled all the requirements of the SOSORT Braced Patients Management Guidelines: answering to the specific questionnaire reported in the paper [15], Excellent Results have been obtained, with 43 out of 44 criteria respected; 1 being not applicable. Accordingly, the patients have been followed by a highly trained team focused on maximizing compliance and minimizing the Quality of Life and psychological impacts of brace treatment [36].

Muller [53] in a recent study questioned the use of a TB in the everyday clinical practice, concluding that it is not to be recommended because patients might feel wary and the mutual trust needed for the confidential patient-doctor relationship might be negatively affected by the use of objective checking tools. According to these premises, and due to the same fear, we were very cautious at the beginning: The TB was presented to patients and family as a useful device with which to optimize the efficacy of therapy, explaining the importance of wearing a brace and of knowing the real wearing time to ensure adequate therapy with respect to the results actually obtained. When receiving patients, all the particulars offered by the software allowed a deepening of the real everyday use, making it possible, for example, to check for less compliant days of the week or for specific periods of not wearing, for daily transitions (brace on/off cycle) and so on. Consequently, we discovered that, when well used, the TB strengthens the patient-physician relationship through a frank discussion of real data, avoiding guesses that are sometimes totally wrong. Those guesses, if wrong, can really undermine a good relationship. Our experience has been that the use of the TB can contribute to the good results of therapy and become an integral part of the treatment.

Given the objective measurement of compliance, a reliable profile of compliers and non-compliers can be established so as to allow tailored treatment protocols and educational and supportive efforts to improve compliance. The use of the TB in the group of patients who had already started brace therapy had, as its main objective, the goal of facilitating the physician’s clinical choices: When a brace is not correctly used it is very difficult to dose the therapy adequately. In this group of patients, probably TB has been better accepted by parents than patients. In any case, the use of the TB didn’t have a negative effect on their compliance; on the contrary, after the use of the TB the referred compliance increased.

We developed an algorithm to cope with the seasonal changes of temperature. However, we must consider that the device may not be applicable to the regions with hot weather, so we did a further analysis of the collected data looking for the different regions of our country: the comparison didn’t show significant differences.

We can’t deny that there is the possibility that in the really worst case the hot weather could allow an overestimation (although little) of real compliance, but it is a very rare event and the fact that we use always statistically processed data decreases even more this risk. So the problem potentially exists, and should be better quantified with experimental studies in the future; on the other hand, at our latitudes (continental climate) the phenomenon appear for a limited period of the year, such that the influence on the data does not appear to affect results. The distribution of air conditioners and the fact that during the hottest hour of the day people in general, and patients wearing braces even more than the others, prefer to stay at home or in cool places, where the temperature is on average lower than outside, obviously contribute to limit this event.