For supplying CPM movement splints after knee or shoulder operations

There are already numerous studies and reviews on the subject of CPM splint therapy in the literature. In addition to controlled and randomized clinical studies on the use of CPM for various indications [4-12], there are also isolated meta-analyses [13, 14] and reviews [15, 16]. Overall, the study situation turns out to be very heterogeneous with regard to the indication, the comparative therapy, the duration of the study, the start of treatment and the endpoint consideration. In addition to this heterogeneity in the study situation, the fact that the studies come from different countries whose care structure cannot easily be transferred to the German health system [9, 17] makes it even more difficult. Methodologically, the studies are limited, among other things, by the fact that complete blinding of everyone involved is not practical. The treatment results are also influenced by many factors (e.g. type of joint injury, surgical technique, comorbidities). One objective of the present study was to describe this heterogeneity, with a focus on the real care situation. A central component of this care situation is the decision of the health insurance companies as to whether the CPM splint treatment will actually be approved after the prescription has been issued. In the sense of “real-world evidence”, this not only corresponds to the actual reality of care, but also does justice to the variety of possible scenarios in which patients can find themselves after knee or shoulder surgery.

>> The studies listed have already provided numerous evidence of the therapeutic benefits of CPM movement splints. As part of its preliminary report in January 2018, the IQWIiG also confirmed a hint or indication of an additional benefit of CPM splint therapy for 3 out of 5 indications [18].
However, the focus of the work published so far has rarely been on the care context of outpatient medical aids in Germany. For a tool that has been used for around 40 years, data on “real-world evidence” is comparatively rare. This is where this publication comes in.
The central objective is to describe the empirical care situation of patients who received a prescription for a CPM movement splint for the postoperative follow-up treatment of their knee or shoulder joint surgery. In addition to the descriptive representation of the empirical care constellation, this is done with reference to the everyday effectiveness of the rehabilitation procedures used and patient-relevant parameters such as pain, functionality, quality of life and stability.
methodology
The present prospective care study examines outpatient follow-up treatment after knee and shoulder operations in the German care context using a four-arm non-interventional study design. Patients who received a prescription for a CPM motion splint following their knee or shoulder joint surgery were included. The subjects were interviewed at two times six weeks apart. The assignment to the CPM or control group was made through the approval or rejection of costs by the relevant health insurance company (verum group = CPM splint treatment; control group = no CPM splint treatment).
The survey forms used for the patient survey consisted of three blocks of questions for both patient groups (knee/shoulder). The first section asked for personal and demographic information about the patient, as well as information about the operation. The second section records information on the patients’ follow-up outpatient treatment. The third and final block collects information on the patient-relevant endpoints of the study, i.e. knee or shoulder functionality and quality of life after knee or shoulder operations. Depending on the indication, this questionnaire section contains the translated and validated items from the Oxford Knee Scores (OKS) [19] or Oxford Shoulder Scores (OSS) [20]. Both scores each ask for 12 items that cover different variables regarding the functionality of the joint and the quality of life of the respondents.
The initial population consisted of all patients for whom prescriptions were received from three service providers of the CPM Association following knee or shoulder surgery between May and October 2017. In this context, it is important to emphasize that the population did not include all patients after knee and shoulder operations in Germany, but only a subpopulation of patients for whom a postoperative aid prescription was available. The statistical analysis of the study results was carried out using the programs SPSS Statistics 24.0, R and Microsoft Excel. In addition to descriptive statistics, inferential statistical methods were also used as part of the quality of life analyses.

Number of cases and examination units/period
In order to achieve a statistical power of at least 90% for carrying out a two-sided T-test, a sample size of 236 patients for the knee indication and 176 patients for the shoulder indication was calculated. As part of the knee indication, a power of over 90% is achieved with a sample size of 150 (experimental group) and 37 (control group) in order to be able to show a clinically relevant and significant (p < 0.05) mean difference of 4.7 points. With a conservatively assumed dropout rate of 10% in the experimental group (30% in the control group), 166 and 53 patients were therefore required (N1 + N2 = 219). For the shoulder indication, 99 patients (experimental group) and 46 patients (control group) achieved a power of 90% to show a clinically relevant and significant (p < 0.05) mean difference of 4.7 points. With a conservatively assumed dropout rate of 10% in the experimental group (and 30% in the control group), 110 and 66 patients are required (N1 + N2 = 176).
According to the calculated case numbers, a total of 243 patients (knee) and 177 patients (shoulder) were included in the care study. The patients were surveyed twice with the same survey instrument, 6 weeks apart. The first visit always took place within 10 days of discharge from the hospital or, in the case of outpatient procedures, within 10 days of the operation. In order to be able to evaluate the entire postoperative follow-up treatment (before the first visit in the patient's home), retrospective information about the operation and a possible inpatient stay were requested. The central inclusion criterion was the postoperative prescription of CPM splint therapy following knee or shoulder joint surgery. The examination period was 145 days (knee) and 132 days (shoulder) between May and October 2017.
As stated in the information on the study situation, there are already numerous clinical studies on the effectiveness of CPM splint treatment. The context of outpatient medical aid provision in Germany is often neglected in the literature. The focus of the study is therefore on how patients with different indications are treated. This “real-world evidence” can provide a variety of information about the reality of care that clinical studies often cannot provide due to the random assignment to the study groups. The health insurance company's decision as to whether the costs of CPM splint therapy will be covered plays a crucial role in the reality of care for patients and can potentially have a massive impact on the subjective quality of life of these patients. To our knowledge, this is the first study in which everyday-relevant functionality and thus quality of life is examined as a function of rehabilitation after orthopedic trauma surgery interventions.
Results
The most important results from the statistical analysis of the patient survey are presented below. The order in which the results are presented is based on the structure of the survey forms. First, the information on demographics and the operation is presented. These statements are then supplemented by the results of outpatient follow-up treatment. Together, these two sections already represent a large part of the reality of care for the patients included in the study. The third part of the presentation of results focuses on parameters of everyday effectiveness and quality of life and, unlike the previous two sections, is not devoted to the question of “how”, but rather “how successfully” the patient in question was treated, according to his or her self-assessment. Table 1 summarizes the results of the first two questionnaire sections. These results will be examined in detail below.

Demographics and operation information
A total of 243 patients (knee) and 177 patients (shoulder) took part in the survey. The control groups were significantly more affected by dropouts between both visits. This circumstance was already expected due to the survey mode and was taken into account in the sample size calculations.
The time interval between the two survey times after knee operations was an average of 42.8 days. For patients after shoulder surgery, this value was comparable to 42.1 days. Both statements correspond to the predetermined examination period of six weeks.
After knee operations, a similar number of subjects of both genders were included in the study. In the CPM group, men provided slightly more information (53%, compared to 46% women1). Both genders were equally represented in the control group. After shoulder surgery, the proportion of women was significantly higher (57% in the CPM group and 69% in the control group).
As mentioned in the introduction, patients after various surgical procedures were included in the study. To ensure comparability between CPM and control groups, these operations were evaluated separately for CPM and control groups. The most common operations in knee surgery are anterior and posterior cruciate ligament procedures, cartilage procedures and major meniscus procedures. This applies to both the CPM and the control group. Clear parallels can also be seen between the most common indications for patients after shoulder surgery. In both groups, operations in the area of impingement/ASD/calcified shoulder/frozen shoulder were performed most frequently. Complex shoulder procedures with multiple indications and rotator cuffs were the second and third most frequently documented.
If you compare the results of the information on demographics and the indications between the study groups, the similarities are particularly striking. Despite the lack of randomization, sufficient comparability between CPM and control groups can therefore be assumed, even without dedicated propensity score matching.
In addition to the type of operation, it was also recorded whether it was an inpatient or outpatient operation. In the former case, the duration of the inpatient stay was also asked. After knee surgery, outpatient operations were documented slightly more frequently in the control group (28%) than in the CPM group (26%). Exactly opposite proportions between the study groups were reported for shoulder operations (CPM group 39%, control group: 26%). The majority of reported operations took place in an inpatient setting across groups and indications. The average length of stay after knee surgery was comparable for the CPM group (3.3 days) and the control group (3.1 days). However, the average hospital stay after shoulder surgery was significantly longer for patients in the control group (3.6 days) than for patients with CPM splint treatment (2.2 days).
The study subjects' employment status and inability to work resulting from the operation or treatment were also asked. In particular, the results on employment status provided relevant information about the supply situation. The proportion of employed people was significantly higher in the CPM group than in the control group for both patients after knee surgery (70% to 56%) and after shoulder surgery (60% to 31%). This can be seen as a clear indication that working patients who want to be able to return to work as quickly as possible are significantly more likely to be treated with CPM splint therapy. The significance of these results is further reinforced by the information on the duration of absences from work due to operations or rehabilitation. After knee surgery, it was found that patients in the CPM group (33.3 days) took slightly less time to return to work than patients in the control group (34.1 days). This connection was even more clearly visible in the context of the shoulder indication. There, patients without CPM splint therapy were on average unable to work for significantly longer (41.3 days) than the splint patients (33.2 days). Information on early retirement was not collected.

Outpatient follow-up treatment
In addition to the CPM splint treatment, physiotherapy is very often part of the follow-up treatment after knee and shoulder operations. Therefore, the examinees were questioned in detail about any physiotherapeutic services during the examination period. The proportion of patients who used physiotherapy services as part of outpatient follow-up treatment was higher in the control group after knee surgery than in the CPM group (78% to 71%). After shoulder surgery, 78% of the patients in the control group were also treated with physiotherapy. This applied even more frequently (82%) to patients in the CPM group. The two groups did not differ significantly from each other in terms of the average number of treatments (approx. 2 applications/week), the average session duration (between 25-30 minutes) and the number of appointments before the first visit. This was not to be expected due to the strict regulatory requirements of the medicinal product guidelines. Depending on the indication and examination group, between 20-30% of all patients reported difficulties in making appointments for physiotherapy (see 1).
During the study, outpatient splint therapy was only used for patients in both CPM groups. The average outpatient splint therapy after knee surgery took just over half an hour (31.72 minutes per application) and was performed by patients almost four times a day (3.86 units). Comparable values have been reported for patients with CPM shoulder therapy. The average application duration was 26.03 minutes, slightly shorter than in knee patients and was used 3.69 times daily.
It was also asked whether other aids were used in addition to CPM splint therapy and/or physiotherapeutic measures. In this context, after knee operations, 86% of the CPM group and 92% of the control group reported that they received additional aids. A separate analysis of open patient information showed that these mostly involved crutches and orthoses or bandages. For example, almost 90% of all patients received a crutch after knee surgery. Other aids were documented less frequently in the context of shoulder operations. Around 64% of the patients with CPM splint treatment and 42% of the patients without CPM splint treatment received additional aids during the follow-up outpatient treatment. Orthoses/bandages and electrotherapy devices were most commonly used in this context.
Physiotherapy and CPM splint treatment are not the only possible therapy options as part of the follow-up treatment after knee and shoulder operations. Lymphatic drainage was most often listed as another therapeutic service in connection with both indications.
Another relevant aspect of the reality of care for outpatient follow-up treatment is the medication consumption of the examinees. It was examined to what extent the different treatment options for pain are supplemented by taking medication. After knee operations, the proportion of patients who had to take painkillers in the CPM group (59%) was significantly below the comparison value in the control group (72%). Something similar could also be observed after shoulder procedures. Here, 69% of the patients in the CPM group and 79% in the control group documented the use of painkillers. The most frequently mentioned active ingredients were diclofenac, ibuprofen and metamizole. In summary, it can be said that CPM splint treatment is associated with significantly lower drug consumption. This should be seen as an indication that patients with splint treatment suffer from less pain compared to patients who were denied this treatment option. As part of the evaluation of the quality of life parameters, the pain parameter is also documented.

Quality of life and functionality
The functionality of the knee and shoulder joints as well as the quality of life of the patients were assessed using 12 questions each from the Oxford Knee Score and Oxford Shoulder Score. For all 12 individual questions, the respondents were asked to provide information on the functionality of their knee joint and their quality of life on a five-point scale (0 – 4). A higher value corresponds to a higher functionality and a higher health-specific quality of life for the patient, while lower values reflect a worse condition of the patient. With 12 questions, a total result of a maximum of 48 points was possible. Table 2 shows the individual results for the 12 items of the OKS. The results of the first visit were comparable in both groups. At the second visit, the results for patients in the CPM group were significantly more positive compared to the control group for all 12 individual items of the OKS.
The evaluation of the functional scores provided highly significant (p < 0.001) results in favor of the CPM group. The mean Oxford Knee Score improved in the CPM group from 17.2 (out of 48 possible points) in the first survey to 28.5 in the second survey, an improvement of 11.3 points. In the control group there was only an improvement from 16.3 to 21.0, i.e. by 4.7 points. This improvement is additional in the two density distributions in Figure 1
visualized.
When the effects were statistically checked, the results of the Oxford Shoulder Score were even clearer and also highly significant (p < 0.001) in favor of the CPM group. The mean improved in the CPM group from 11.6 (out of 48 possible points) in the first survey to 25.5 in the second survey, which corresponds to an improvement of 13.9 points. In the control group there was only an improvement of 5.2 points from 10.9 to 16.1 points. Finally, it should be noted that across both indications it was clear that outpatient splint treatment had a significantly positive effect on the joint functionality and quality of life of the examinees, which remained particularly true when controlling for the third variable “physiotherapeutic treatment”. Particularly noteworthy are the results of the questions about pain in connection with the aforementioned lower consumption of painkillers by patients in the CPM group.
In addition to the tabular representation (Table 3), the overall OSS values were also visualized using the density distributions for both visits (Fig. 2). As with the knee indication, it can be seen in this figure that both distributions overlap more at the first examination time, while the distribution of the CPM group at the second visit clearly shifts towards a higher quality of life for the patients.
The differences in the distributions between CPM and control groups become even clearer when looking at the box plots in Figures 3 and 4. The increase in quality of life and joint functionality between both visits is significantly more visible for patients with CPM splint therapy than for patients in the control groups.
Discussion/limitation
The results provide information about what the reality of outpatient follow-up treatment after orthopedic surgery looks like in Germany. In addition to the results on quality of life parameters, it is particularly the patient information on outpatient follow-up treatment that represents a valuable addition to the previous studies.
The proportions of patients who received physiotherapy treatments, CPM splint therapy or other therapy services clearly show that the first two treatment approaches mentioned in particular have become established in healthcare practice. Since only patients who were prescribed CPM splint treatment were included in the study, this applies in particular to physiotherapy, which was almost mandatory in the context of knee and shoulder operations.
The addition of information about the difficulties in finding an appointment and getting to the physiotherapeutic treatment center to the proportion of patients receiving physiotherapeutic treatment also represents an important part of the reality of care. Such details about the actual care situation were systematically excluded in the often clinical studies carried out to date [6, 7, 11] or did not refer to outpatient follow-up treatment but to inpatient follow-up treatment [21].
There are a few other comparable studies on the care situation in Germany that include multiple indications, but they primarily deal with the comparison of splint therapy versus physiotherapy [4, 22]. However, in accordance with the specifications of the list of medical aids [3], the results of the present care study underline that the combination therapy of physiotherapy and additional CPM movement splint is a common and promising treatment option.
What is particularly striking is the low proportion of other therapy services (between 4-21%, depending on the examination group and indication).
Combining data on outpatient follow-up treatment (particularly medication consumption) with the Oxford quality of life and functionality scores clearly showed that patients with CPM splint therapy not only consistently reported less pain, but also required pain medication less frequently. There were statistically significant advantages for patients who were treated with CPM splint treatment both in the pain parameter and in all other parameters relating to quality of life and functionality. This expands the recently published results of IQWiG, in which there was an indication of a greater benefit of CPM splint treatment for the indication of frozen shoulder with regard to the endpoint pain [18], to include further indications.
Overall, the results on quality of life are also worth highlighting because IQWiG confirmed a need for additional quality of life data in its preliminary report on CPM splint therapy [18].
In this care study, the focus was for the first time on a systematic description of the reality of care. The additionally collected parameters of the Oxford scores clearly indicate that CPM splint therapy has a positive influence on the patients' quality of life and joint functionality.
A limitation of this study is the selection bias, which is due to two methodological decisions within the health care study. The first selection bias results from the selected starting population. Since only patients for whom CPM splint treatment was prescribed were included, no statements can be made about those patients in whom the treating physician saw no need for splint therapy.
The second selection bias relates to the indication. The aim of covering as many indications as possible in one study design is also accompanied by the risk of distortions between the groups. This circumstance was significantly attenuated by the documentation of the types of operations by the patients (and the extensive agreement on these), but should be listed as a limiting factor.
In addition to information about the patient's employment status, information about possible early retirement could also be important. This information was not generated as part of the data collection. This leaves room for further studies that, for example, look at the effects of knee and shoulder operations on social productivity losses. <<