What to Expect: Medical Quality Outcomes and Achievements of a Multidisciplinary Inpatient Musculoskeletal System Rehabilitation What to Expect: Medical Quality Outcomes and Achievements of a Multidisciplinary Inpatient Musculoskeletal System Rehabilitation

The incidence of chronic diseases is rising. Rehabilitation plays a vital role in preventing and minimizing the functional limitations associated with chronic conditions and aging. Routine outcome measures include disease-specific and unspecific general health param- eters. This study evaluates indicators for medical quality outcomes from 10,373 patients (61.00 (cid:1) 13.65 years, 51.7% women) who have undergone orthopedic rehabilitation for three weeks. Inpatient rehabilitation reduces lifestyle-related risk factors, optimizes organ functioning and improves the well-being in the majority of patients (81.3%; SMD = 0.52 (cid:1) 0.38). Improvements of unspecific and indication specific outcome parameters can be observed in a comparable magnitude. However, disease specific and unspecific health factors are not directly related to each other (r = 0.19). Age, gender, ICD-classification and time of rehabilitation have an influence on initial values and on indication-specific medical outcomes but are insignificant with regards to improvements in unspecific medical out- come parameters. Inpatient rehabilitation includes two main pathways of medical practice, which can be clearly distinguished in terms of their therapeutic outcome. There are general health interventions, such as lifestyle modifications, diet and physical exercise, and symptom-specific treatments. So multidisciplinary medical rehabilitation improves general well-being and physical functioning as well as reduces risk factors in the majority of patients. the federal Austrian social security institutions involved in quantifying the quality of rehabilitation outcomes are the basis of this work. We provide descriptive standardized numeric indicators of the rehabilitation process as well as monocentric reference data for a 3-week inpatient orthopedic rehabilitation program. Data collection was performed by doctors and healthcare professionals during routine medical treatment. Standardized clinical characteristics of patients were recorded systematically at the time of admission and discharge. The easily quantifiable medical parameters included general health characteristics such as body measurements and


Introduction
The definition of rehabilitation has become more transparent all over the world after the World Health Organization (WHO) published the five health strategies (promotion, prevention, cure, rehabilitation and palliative care) in rehabilitation [1]. Rehabilitation was previously understood as a heterogeneous approach. It was framed as a highly specialized service for athletes, a controlled training intervention or a post-injury service for return-to-work intervention. The definition was standardized through the WHO to a homogenous understanding of rehabilitation as a service to restore functioning, ameliorate the impact of the reduction in capacity and minimize further initial health problems in all stages of health provision [2].
Special care in rehabilitation is given to the ailments of the modern society. As such, especially risk factors like sedentary behavior 1 , physical inactivity, overeating causing obesity, alcohol abuse and smoking leading to various chronic diseases that immediately affect the musculoskeletal system [3] are causes for concern. Due to those lifestyle choices that lead to potentially life-threatening conditions, physical activity has become an important therapeutic approach. It is a primary, secondary or tertiary therapeutic approach of chronic diseases over the last decades. Physical activity recommendations are now broadly understood as health-related interventions that propose a minimum dose of physical activity.
The WHO [1] suggests at least 150 minutes of moderate aerobic activity or 75 minutes of vigorous aerobic activity throughout the week for adults. In addition, muscle-strengthening activities should be done involving major muscle groups on two or more days a week and sedentary behavior should be reduced. Small amounts of physical activity already have a positive impact on the health status; however, the strongest effect can be observed in adults that transformed from an inactive to an active lifestyle [4][5][6].
Medical intervention programs therefore do not only aim to restore physiological functioning, reverse or undo the damage caused by disease or injury, but also rather to optimize the health status by improving strength and aerobic capacity. The amelioration of the physical constitution leads to health-promoting effects like the reduction of high blood pressure, a better glucose profile and reduced blood lipids [7,8]. Physical activity has a positive impact not only on metabolic and cardiovascular diseases but also on the musculoskeletal system.
Complaints of the musculoskeletal system are most commonly those of inflammatory and degenerative origin. These origins of disease are the cause for chronic pain, painful functional impairment and a reduced quality of life worldwide. Rheumatoid arthritis is found in 3.2% of women and 1.9% of men. Furthermore, another increasing ailment is osteoporosis, which is found in 3.1% of women and 1.9% of men, with a high prevalence from the age of 80 onwards [9]. 1 For example, through (screen) workplaces and automation, leisure activities such as video games, social media and television.
The most common symptom of the musculoskeletal system is back pain (60-80% long-time prevalence [10,11]). The prevalence is dependent on age, with a rise of prevalence with age. The highest prevalence for back pain is set between 40 and 69 years; women are more affected than men [12,13].
The findings of an analysis by the Austrian Health Survey 2014 [14] report an estimated rate of 24% (about 1.8 Mio individuals) suffering from chronic back pain. Furthermore, 19% claimed about discomfort of the cervical spine. The cervical spine complaint was gender dependent (23% women versus 14% men). Arthrosis was found in 15% of women and 8% of men. The gender difference refers not only to the frequency of the disease but also to the pain perception, due to a higher rate of female pain sensitivity, lower threshold for pain and a lower suspension of pain [15].
Due to the high prevalence of degenerative diseases of the musculoskeletal system, there is also a demand for surgical treatment with endoprosthesis material. The highest operative treatment rates for such procedures worldwide are found in Switzerland, Germany and Austria. In Germany, about 230,000 hip, 170,000 knee and 25,000 shoulder endoprosthesis surgeries were performed in the year 2011 alone [16].
Physiotherapy and physical activity are fundamental components for the process of recovery. The body of available knowledge confirms positive effects of physiotherapy and physical activity throughout all organ systems. It reduces swelling as well as promotes building strength and aerobic capacity. Through the process of physiotherapy and physical activity interventions, a total physiological functionality after surgical treatment can be achieved. Moderate physical activity reduces the risk of osteoporosis and improves the osseointegration of bone substitutes. Almost all musculoskeletal diseases can be treated by therapies including physical activity and strengthening programs [17][18][19][20]. Physical activity thus has become the gold standard among medical treatments.

Orthopedic rehabilitation
Following the WHO definition, the rehabilitative process can be divided into four phases: phase I includes the early mobilization in primary treatment, phase II provides follow-up treatments or post-acute therapy in rehabilitation centers, phase III tries to integrate and stabilize the long-term life modifications and phase IV deals with long-term rehabilitation including a probably outpatient aftercare. The aim of orthopedic inpatient rehabilitation (phase II) is to restore the health status based on the bio-psycho-social health model.

Situation in Austria
Musculoskeletal diseases, caused by inflammatory, degenerative processes or injuries, permit an inpatient orthopedic rehabilitation (WHO, Phase II [21,22]) over 3-4 weeks for restoring physiological functioning and reintegrating to social and professional life.
Based on historical decisions [23] and Health Technology Assessments (HTA) (e.g., [24]), there is currently a framework of contracts with the federal Austrian social security institutions entities in place 2 . This includes performance agreements for orthopedic rehabilitation (see [25]), based on criteria regarding the quality of processes and treatment outcomes [21]. Criteria include individual detailed results that demand a standardized statistical recording of outcome parameters from admission up to discharge of the rehabilitation program. Furthermore, evaluation and statistical analyses of these medical outcomes are generally not open to the public [26].

Implementation of orthopedic rehabilitation in Austria
There is strong evidence that rehabilitation is necessary as part of the treatments for inflammatory or degenerative diseases, as well as for postoperative conditions or injuries of the musculoskeletal system. The underlying condition's severity and expectancy of the restoration of physical function are requirements for obtaining an inpatient orthopedic rehabilitation. For patients affected by musculoskeletal disease, medical treatment and a large range of rehabilitative treatments are available. The inpatient treatment lasts on average 2-3 hours per day. An individual rehabilitative program consisting of active and passive treatments is provided (see Table 1). Active treatments consist of physical activity including gymnastic and individual physiotherapy sessions and the medical training focuses on underwater , ergometer , Nordic walking , strength , balance , relaxation and motion training. Passive treatments contain sessions like massages, thermotherapy, electrotherapy, ultrasound and educational lessons like various lectures, or psychological coaching. Each patient is offered a program of at least 1800 therapy minutes during 3 weeks, split up into approximately 50% active and 50% passive treatments that highly surpass the physical activity guidelines of the WHO (150 minutes workout in moderate intensity and strengthening exercise twice a week). In particular the pension insurance institution.

Physical Therapy Effectiveness
Medical rehabilitation is structured in different ways all over the world, although a tendency for a standardization of the social and health system can be observed. In Germany and Austria, the social insurance offers inpatient orthopedic rehabilitation over a period of 3-4 weeks in specialized rehabilitation centers with an interdisciplinary team. Some other countries only provide outpatient rehabilitation [27].

Efficacy and sustainability of orthopedic rehabilitation
Diseases demand a certain continuity of the therapeutic process for success after rehabilitation. This means that rehabilitative programs should focus not only on physical activity programs during the stay but also on sustainability of physical activity after the rehabilitation. Therefore, an inpatient stay should also lead to a health-related modification in lifestyle. The efficacy of inpatient rehabilitative programs and especially their sustainability over a long period are important research issues. Literature shows that during the rehabilitative stay, pain can be significantly reduced, and a long-term improvement of physiological functionality can be achieved. The success can be measured even after 12-21 months after rehabilitation [28,29].
The fact remains that physical activity programs have a positive impact on the physiological performance of patients. However, the dropout rate for long-life active and health-related lifestyle change maintenance is high. Findings support the sustained efficacy of an intense, multimodal orthopedic rehabilitation, with moderate evidence, including an improved subjective health status and reduced pain [30]. For rehabilitation treatment after hip and knee endoprosthesis (TEP) [31][32][33][34] and chronic back pain [35][36][37], a level of evidence Ia/Ib is stated. The combination of physiological and psychological training significantly leads to positive changes [18,38,39]. The data confirm that a combination of physical activity and psychosocial training-based treatments has an influence on pain reduction and mental well-being.
Studies in other countries state that about 57% of patients in Europe and even 70% of patients in the US are able to restart regular full-time work and show significant improvements of psychosocial, physiological parameters after inpatient rehabilitation [40][41][42][43]. An interdisciplinary treatment of medical, physical activity based and psychological therapy ensures a high return to professional life [44,45].

Methods
The medical outcome parameters required by the performance profile of the federal Austrian social security institutions involved in quantifying the quality of rehabilitation outcomes are the basis of this work. We provide descriptive standardized numeric indicators of the rehabilitation process as well as monocentric reference data for a 3-week inpatient orthopedic rehabilitation program. Data collection was performed by doctors and healthcare professionals during routine medical treatment. Standardized clinical characteristics of patients were recorded systematically at the time of admission and discharge. The easily quantifiable medical parameters included general health characteristics such as body measurements and cardiovascular parameters, psychological indicators such as pain and subjective health, as well as specific indicators such as daily activities, motor function and physical performance (see Table 2).
These quality-of-outcome measures are to be documented in the discharge report at the beginning and the end of the phase II rehabilitation (this correlates with the methodology of a pre-post comparison). Respectively, data were also summarized using a descriptive evaluation in terms of content and factor analysis. This provides us a simple descriptive evaluation model through independent factors for the sake of a complete evaluation of the outcome quality [26].

Orthopedic reference sample at the clinical trial center
During the service period from 2016 to 2018, a total of 10,373 patients (61.00 AE 13.65 years, 51.7% women) were enrolled in a specialized interdisciplinary hospital for rehabilitation (Humanomed Center, Austria) ( Table 3). A categorization of reasons for hospital treatment was based on the admission diagnosis. Reasons for admission were musculoskeletal diseases (M, 85%) or injuries (S, 15%). Categorizations were based on ICD-10. The majority of patients were admitted for rehabilitation due to back pain (29.6%) followed by upper/lower limb joint as well as knee (21.4%), hip (16.3%) and shoulder (12.0%) disorders. About 73.3% of the patients were admitted after surgery (OP) 3 . Just under a half of those patients (42.2%) were treated with an endoprosthesis (EP) (overall, 31.9%). The proportion of EP rehabilitation procedures was 64.7% for knee patients and 84.7% for hip patients. The period between acute care rehabilitation (OP) averaged 8.4 AE 4.0 weeks. However, patients with knee and hip TEP progressed from phase I to phase II earlier (median, 6 weeks) 4 .
The average inpatient length of stay was 22.00 AE 2.6 days; 2.2% of all patients discontinued their inpatient treatment prematurely due to a loss of rehabilitation capacity (for example, acute illness) or for private reasons (criterion ≤ 18 treatment days). In 3.8% of the patients, the inpatient stay was 4 weeks.

Ethical aspects
This study (A retrospective study-Routine Outcome Parameters of an Inpatient Rehabilitation in Austria) was reviewed and approved by an Ethics Committee ( *Standard scores, also called z-values (z), are calculated by subtracting the (population) mean from an individual raw score and then dividing the difference by the (population) standard deviation (here: for all orthopedic patients). As shown in the results section, deviations from zero respectively mean differences of z-values can be interpreted as an effect size (c.f. Cohen's d).  Table 3. Sample number of patients undergoing rehabilitation of the musculoskeletal system. 9, 9020 Klagenfurt. The data from the hospital information system upon which the publication is based were compiled in compliance with all regulations of the Austrian Privacy Act and the Declaration of Helsinki in its current version. Data collection was done also in accordance with national legalization (hospital statutes, contracts with insurance-authorized institutions, etc.) through the rational of mainly collecting scientific data in the public interest.

Medical outcome quality
Medical outcome quality is defined as the "measurable change in the profession-ally assessed state of health, the quality of life and the satisfaction of a patient" (see Austrian Federal Quality of Healthcare Act, GQG). The outcomes become visible by "the difference between the initial state and the state at treatment end" (c.f. [46]).
The outcome measurement in rehabilitation can be based on various methodological approaches, such as questionnaires, performance tests, equipment measurements and functional physical examinations. The outcome quality measurement, "the outcome," includes features on health (e.g., symptoms and pain), functional levels (e.g., performance) and educational levels. In addition to the patient's subjective assessment ("patient-reported outcomes (PRO)"), medical, diagnostic and other relevant outcome measures/criteria (e.g., ICF) are documented by healthcare professionals.
The aim of the present work is to provide a valid basis for routinely assessing the quality of medical outcomes (routine outcome measurement) based on common data acquisition. The focus therefore is on general (nonspecific, body constitution based) and indication-specific parameters. Continuously collected medical reference values have diverse potential benefits in quality assurance, awareness raising, goal setting and decision-making. The data are also important for evaluation of different care models.

Quantifying "medical outcome quality"
In addition to the descriptive analysis of single indicators, the analysis evaluates the effects of inpatient rehabilitation stay and the medical quality outcome (MQO). This is calculated on the basis of representative monocentric normative data (N = 10,373; see Table 3). The selection of clinical parameters follows the requirements of federal social security instate in the performance profiles of accredited Austrian institutions for outcome analysis, which should guarantee a comparable medical service quality standard. A success index is calculated by summarizing the compulsory basic clinical data of the patients (see Table 2 and [26]).
1. The "Unspecific Health Index" is the arithmetic mean of three independent areas of measurement: body measurements (BMI and abdominal circumference), cardiovascular parameters (blood pressure and resting heart rate), discomfort (visual analogue scales (VAS pain, [47]) and subjective health status (EQ-VAS, [48])). These parameters should provide a simple and quick overview of "unspecific" effectiveness of a rehabilitative stay.

2.
The "Specific Ortho Index" corresponds to a z-normalized mean of "activities of daily living" (EQ5D), "function" (Roland-Morris, WOMAC or Constant-Murley, depending on the affected body region) and the "physical ability" (walking tests).
The Unspecific Health Index can be interpreted as an indicator of general health status or the current body constitution, whereas the Specific Ortho Index corresponds to a disease-specific health parameter (indication-specific outcome). Both indexes together form the "medical quality outcome" in equal parts.

Statistical methods
Based on the value distributions, the individual outcome parameters were transformed into z-values, which allowed a conversion into percentiles. By means of the z-standardization, different scaled quantities can be summarized and changes can be uniformly quantified Individual missing values were not replaced for statistical analysis. The specification of pvalues was omitted, instead effect sizes were used [partial Eta 2 and standardized mean differences (Cohen's d or z-values/-differences, SMD)]. The representation of the partial Eta 2 (η 2 ) was chosen because even the very small numerical differences became statistically significant even if they were not relevant in terms of content and clinical relevance. A η 2 between 0.01 and 0.06 corresponds to a small effect. Occurrences of 0.06-0.14 correspond to a middle effect and values >0.14 to a large effect [49]. The application of multivariate variance analyses (MANOVA) and a factor-analytical reduction of the basic clinical data to individual factors or a total value (MQO) follow substantive, statistical considerations for better clarity.

Results
The effect of the rehabilitation stay and the change between the initial state and the discharge state differ between the individual success factors for medical rehabilitation (see Tables 4, 6 and 7). Specific and unspecific outcomes show a comparable change in sensitivity (η 2 unspecific = 0.522 [UH i ] vs. η 2 specific = 0.540 [SO i ]). The relationship between specific and unspecific outcome characteristics (changes) is small (r < 0.20, see Figure 1 and Table 5), which is reflected in a multivariate analysis of temporal changes in the variance (η 2 unifactorial = 0.654 The nonspecific overall score-the "Health Index (UH Index )"-shows that 72.5% of patients benefit directly from the rehabilitation stay. About 21.8% of the patients remain unchanged and 5.8% worsen between the beginning and the end of the rehabilitation (see Table 4). The detailed analysis of average improvement in general health features by 13.58 percentile points shows that anatomical features, such as body mass index and abdominal circumference, remain unchanged over the 3 weeks in the majority (86.4%). By contrast, cardiovascular features such as blood pressure and resting heart rate are directly influenced by inpatient rehabilitation, with an average improvement of 11.5 percentile points. The most pronounced effects of inpatient rehabilitation can be seen in the symptoms of complaints, where almost every rehabilitation (84.9%) reported a significant improvement. A similarly positive change can be seen in the ADL score (+17.6 percentile points), motor function (+22.2 percentile points) and physical performance, where 2/3 (67.5%) of patients can improve markedly. Almost all considered medical factors (MED, ORT) provide a comparable contribution to the overall success (MQO, all r ≥ 0.40, except for MED1: r = 0.11). The correlation between Unspecific Health Index and Specific Ortho Index is relatively low (r = 0.19; see Figure 1 and Table 5). Therefore, immediate changes in unspecific health scores are not directly associated with improvements in (indication-) specific functional characteristics. One exception is the (subjective) complaints (MED 3), which are related to specific outcome changes (r = .34).
However, the extent of improvement in medical outcome and the effect size are similar in all areas (see Tables 4, 6 and 7 and Figure 1). Just the anatomy factor (MED 1) changed only slightly during rehabilitation (η 2 MED1 = 0.072; Tables 6 and 7). The individual parameters for all patients are presented in Tables 6 and 7 below.

Output values and descriptive data for individual measurements
The starting point (initial) values of rehabilitation clearly show the deficits of the affected patients (see Tables 6 and 7). The average BMI is 28.9 AE 5.4 units (37.0% of patients have a BMI > 30) and 77.3% of patients have "high-normal" or "hypertonic" blood pressure. The perceived pain (VAS; 0-10) of patients is 3.8 AE 2.1 and the subjective health status (EQ-VAS; 0-100) is estimated to be 63.7 AE 16.6% on average.  Bold values show the correlation of each health parameter with the overall medical quality outcome. Table 5.
Correlations between changes in MQO factors. Bold values show summarized factors of outcome measures from unspecific general health parameters (see Table 2). Table 6.

MQO indicators & factors
Starting point values and changes in individually measures quantities. Bold values show summarized factors of outcome measures from disease-specific health parameters (see Table 2). Table 7. Initial values and changes in individual measurements.
Based on the subsample survey of knee and hip TEP patients, it can be seen that changes occur in the nonspecific factors independent of the time of rehabilitation (interaction: time x post-op week: η 2 = 0.001). This result is in contrast to (disease-specific) outcome characteristics (SOi) where the time of onset (post-op week) plays a more important role (interaction: η 2 = 0.061; see Table 9 (cf. [26]).

Comparison of outcomes based on ICD-classification
If one considers the initial medical evaluation values and changes there of as a result of the inpatient rehabilitation stay, we need to evaluate also the initial severity and admission diagnosis. The admission diagnosis was evaluated based on the standardized ICD-10 classification ("ICD") (see Tables 8 and 9). This classification shows that knee patients have worse MED 1 (Anatomy) values at the beginning of rehabilitation (0.36 AE 0.96). In back pain patients, especially negative MED 3 initial medical evaluation values (Discomfort) are prominent (MED 3: η 2 = 0.068). The symptom-specific characteristics (ORT, SO i ) and their differences in initial values were somewhat less pronounced (SO i : η 2 = 0.010 vs. UH i : η 2 = 0.017; see Table 9).
Overall adding up the MQO factors, the initial medical evaluation values are comparable between the ICD diagnostic groups (η 2 MQOi = 0.004), but in individual cases, they certainly play an important role (η 2 multivariat = 0.080, see Table 9).

Influencing factors of age, gender and initial values
Depending on further grouping characteristics (between factors), it is shown that gender (η 2 multivariat = 0.076) and age (η 2 multivariat = 0.067) contribute a significant amount to initial values, which is lower within unspecific than in specific parameters (see Table 9; (cf. [26]). Additionally, medical initial values are influenced by the factor injuries (S) vs. chronic conditions (M; η 2 multivariat = 0.058), as well as the symptom presentation "ICD" (η 2 multivariat = 0.080).
Potential success through rehabilitation with unspecific outcome indicators (changes of condition pre and post, see also Table 4) can be observed within all grouping characteristics of a similar magnitude (interaction time x between factor: η 2 UHi < 0.003 see Table 9). In contrast, age, ICD, postOP and initial value of MQO play a role for alterations in specific outcome indicators (Age: η 2 OHi = 0.039, ICD: η 2 OHi = 0.040, postOP: η 2 OHi = 0.061 and initial MQO value: η 2 OHi = 0.143). All these factors influence the specific outcome significantly. Older patients (>61 years of age), patients with knee and hip issues, patients who enter phase 2 earlier after surgery (<6 weeks) and particularly patients with worse medical initial values show a more advantageous rehabilitation outcome for symptom-specific indicators (not shown).

MQO indicators & factors
Initial values (beginning of rehabilitation; prae) Change from admission (prae) to discharge (post) Classification according to ICD-10 (cf. Table 3 Bold values show overall means of summarized medical quality factors for the three most common orthopedic diagnoses. Table 8. *Initial state at the beginning of rehabilitation. **Differences (improvements) from admission to discharge (corresponds to the interaction: time x factor).
Bold values show the influence (effect size) of the approval diagnosis on baseline values and the impact of other moderating variables on summarized unspecific and specific health factors. Table 9. MQO and potential influence factors.

Discussion
In times of dwindling resources, evidence-based justifications of medical measure effectiveness are gaining ever more practical and healthcare economic importance. New possibilities of medical treatment measures and constantly changing framework conditions are a challenge for quality assurance management, which uses structural, process and outcome characteristics to evaluate the degree to which predefined goals have been achieved. Potential environmental influences or a change of framework conditions can also be evaluated with regard to the MQO. The recorded routine data are used both as an evaluation criterion and as comparison data of the expected MQO or output values.
Current treatment paths emphasize a stratified approach, active therapies and educational measures, whereby evidence of effectiveness and clinical practice still clearly diverge [28,50,51].
Based on presented clinical observations through routine operating procedures of a multidisciplinary inpatient musculoskeletal system rehabilitation, evaluation can be performed. The evaluation takes into account facility and patient comparisons and is based on standardized assessments of different treatment options as well as factors of outcome quality. Existing performance profiles have a certain amount of latitude to focus on treatment or to apply promising therapeutic options. Such differentiated treatment pathways are necessary for efficient and successful treatment so that individual patients or specific patient groups can be addressed.
A physician/therapist or facility can use the results presented to make an evidence-based decision relatively quickly on whether the treatment process is proceeding to their satisfaction. Orthopedic patients receive a realistic assessment of what and how much they will improve their physical abilities through the 3-week inpatient medical treatment.
The selection of outcome parameters follows evidence-based and economic considerations that should guarantee a comparable standard of quality medical treatment. Indication-specific characteristics (ORT, SO i ) are at the center of the rehabilitative treatment, to restore the ability to function and work or to reintegrate into the social and professional environment. Another focus is on nonspecific health features such as individuals that are overweight, have high blood pressure or are physically inactive. These characteristics are associated with poorer health, cardiovascular disease and metabolic disorders. They are among the most important variable risk factors for chronic diseases and premature death [52].
An important task in inpatient rehabilitation is, in addition to the individual symptomatic treatment of a patient, to sustainably reduce these risk factors. Quality of life and functioning are characterized by positive lifestyle modifications, for example, an increase in physical activity. A reduction in the recorded basic clinical parameters such as BMI, abdominal circumference, blood pressure, heart rate and pain [53][54][55][56] is therefore also highly relevant in the inpatient rehabilitation of degenerative and inflammatory disorders of the musculoskeletal system [57,58]. A mathematical comparison with "healthy" patient reference data underlines the importance of the MQO factors presented in this work, which deviate in the order of about one standard deviation of healthy individuals.
In order to work meaningfully and deductively, it is helpful to reduce the abundance of partially redundant information to a manageable, uniform level. An additive summary of independent medical areas (factors) to a key figure (UH i or SO i and overall MQO) is to be discussed in terms of content and statistics. These key figures give a simple and quick overview of the "unspecific" effectiveness of the rehabilitative stay for certain healthcare teams or treatment programs.

Nonspecific quality of outcomes
When assessing the effectiveness of inpatient rehabilitation based on the nonspecific health outcome characteristics, it is not so important to use grouping characteristics such as symptoms. Grouping based on gender and age is here too of minor importance. Comparing the values at discharge with the initial medical evaluation data shows that 72.5% of patients benefit directly from inpatient rehabilitation (see Table 4). Significant interactions with grouping features are present but can be classified as "small" compared to the main effects (η 2 < 0.003; see Table 9). The observed strong rehabilitation effect is similar for all subgroups (η 2 > 0.500; see Tables 6 and 8). This unspecific success is probably attributable to all inpatient rehabilitation stays and to the preventive effect of activity or movement in the rehabilitation setting.
The change in the nonspecific MQO is independent of the time of rehabilitation, the post-OP week (η 2 UHi < 0.001; see Table 9). In an untreated "real" control group without rehabilitation, therefore, no positive change in the constitutional state of health is to be expected.
In contrast, the importance of early inpatient rehabilitation as early as possible is revealed in disease-specific outcome characteristics such as function or physical ability (η 2 OHi = 0.061). This underlines the importance of a multidimensional view of specific and nonspecific outcome quality, describing two independent (active) components of rehabilitation (see Figure 1 and Table 5).

Specific quality of outcomes
The most pronounced effects of inpatient rehabilitation are seen in the complaint (MED 3) and indication-specific characteristics (ORT 2, ORT 3 and SO i ). More than 2/3 of patients experience significant improvement of symptoms and specific characteristics (see Table 4). About 77.0% of patients show indication-specific improvement of outcomes in inpatient rehabilitation. However, 19.0% did not show significant changes and 4.0% showed worsening from rehabilitation beginning to rehabilitation end. Therefore, it is important to remember that not all patients can directly benefit from treatment. In sum, (overall MQO) improvement is achieved, in the vast majority (81.3%, SMD = 0.52 AE 0.38; see Figure 1). Improvement of unspecific and indication-specific outcome parameters is achieved in a comparable magnitude.
Unlike nonspecific health scores, moderating factors play a more important role in the specific outcome quality. In particular, patients who enter Phase II earlier (<6 weeks) after surgery and patients with worse initial medical evaluation show better rehabilitation success in symptomspecific characteristics (see Table 9).

The applicability of our results in rehabilitative clinical practice
Despite large international differences in healthcare teams and treatment measures in rehabilitation, the observed effect sizes can support individual evaluation but cannot replace it. In addition to case by case evaluation, it is important to consider the medical focus, the rehabilitative practice and performance profiles, as documented outcomes come about through different treatment programs and the associated different use of healthcare resources.
Absolute values and individual profiles of the MQO should always be evaluated according to the given setting and at the doctor's discretion. Single measurements are subject to a variety of moderating influences and measurement errors. The presented continuous measures of MQO have advantages in terms of their (scale) properties and sensitivity to frequently applied categorical criteria.
When using the presented clinical reference values of the MQO for orthopedic rehabilitation (WHO phase II) in Austria, apart from the reference sample, the monocentric character of the work has to be considered. The need for adjusting for different facility comparisons cannot be definitively answered. Different individual starting values must always be taken into account, as worse outcome parameters at the beginning of rehabilitation are accompanied by a greater potential for improvement (e.g., r MQO = À0.461). Due to the prescribed performance profiles and the centrally controlled assignment modalities by the insurers, however, it can be assumed that the initial values presented and especially the changes in the MQO are representative of the inpatient rehabilitation of the musculoskeletal system in Austria.
The presented results and experiences of the clinical trial center suggest that the expected rehabilitation effects in the MQO factors are universal. Risk adjustment or indication-and group-specific modeling does not seem necessary. A closer characterization of nonresponders and types is still pending.

Limitation
The practical significance of the MQO outcomes or relationships with external criteria (endpoints), such as the incapacity for work, remains to be tested. Especially the global evaluation between MQO and sociomedical relevant external criteria is a potential further route for development. It is to be clarified which of the observed changes in MQO have significance for the sustainability of rehabilitative measures and to what extent optimized treatment pathways can influence them. Improvements in one outcome may well be accompanied by deterioration in other outcomes. Therefore, multidimensional approaches to the quality of results are always the methods of choice.

Conclusions
In addition to the usual "primary patient-oriented" assessment of quality of outcomes, the subjective assessment and satisfaction through self-reporting of patients as well as quality registers on surgical frequencies, length of stay and complications, these factors present a valuable addition to the medical outcome quality evaluation. These can be essential for decision-making or can contribute to the design processes and further developments of quality assurance in rehabilitation facilities.