The increase of psychosomatic disorders due to cultural changes requires enhanced therapeutic models. This study investigated a salutogenetic treatment concept for inpatient psychosomatic treatment, based on data from more than 11000 patients of a psychosomatic clinic in Germany. The clinic aims at supporting patients’ health improvement by fostering values such as humanity, community, and mindfulness. Most of patients found these values realized in the clinical environment. Self-assessment questionnaires addressing physical and mental health as well as symptom ratings were available for analysis of pre-post-treatment effects and long-term stability using one-year follow-up data, as well as for a comparison with other clinics. With respect to different diagnoses, symptoms improved in self-ratings with average effect sizes between 0.60 and 0.98. About 80% of positive changes could be sustained as determined in a 1-year follow-up survey. Patients with a lower concordance with the values of the clinic showed less health improvement. Compared to 14 other German psychosomatic clinics, the investigated treatment concept resulted in slightly higher decrease in symptoms (e.g., depression scale) and a higher self-rated mental and physical improvement in health. The data suggest that a successfully implemented salutogenetic clinical treatment concept not only has positive influence on treatment effects but also provides long-term stability.
Since 1948, the World Health Organization defines health as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity” [
Over the last century, a major shift in the distribution of disease types has occurred. While in the last century the health system had to cope predominantly with infectious diseases [
The SOC concept was introduced by Antonovsky and encompasses comprehensibility, manageability, and meaningfulness [
Schumacher et al. [
Höfer and Straus [
Antonovsky [
Another extended health model is the biopsychosocial model (BPS) proposed by Engel. The BPS integrates biological information from the patient’s body, psychological aspects such as negative thinking or lack of self-control, and social factors such as socioeconomic status and religion to assess the causes of illness, aiming to bridge the mind-body connection [
Another construct that can be included into the concept of salutogenesis is mindfulness. The most common psychological definitions of mindfulness mainly describe the construct by emphasizing two points: firstly, full attention to the present moment on immediate experience. Secondly, this awareness should be experienced in a state of equanimity, whatever sensation arises, without judgment, elaboration and reaction [
In the present study we intend to show the treatment effects of a psychosomatic clinic which aims to implement the above-mentioned salutogenetic factors. We will investigate how these factors and values are realized in the treatment, and analyze the influence of personal concordance on treatment success.
A positive correlation between spirituality, for example, feeling centered, and therapeutic success in HF has been reported previously [
This field comprises mainly the following: general assembly (“Plenum”) on parting and for welcoming new patients (also open to visitors); assembly of all patients; therapeutic group for support and for improvement of communicative skills; sharing of community tasks and responsibilities; psychoeducation in fields like health and disease, treatment concepts, and healthy nutrition; recreation offers for the weekend; involvement of patients in the improvement and complaint management.
The mental-spiritual aspect of human existence is addressed through various optional offers. Spirituality is actively supported by a systematical introduction and daily exercises of mindfulness meditation, daily changing meditation sessions (guided, silent, moving, contemplative observing the sense of words and useful phrases or basic questions of human life and death, and visualization), the option to practice in a separate meditation room all day, daily self-reflection, autogenic training, and other relaxation techniques.
In this context it is important to mention the experimentally confirmed motivational concordance theory by Hyland et al. [
We will assess treatment effects, also dependent on the respective diagnosis or treatment path, and also analyze the effect of personal concordance on treatment success and sustainability of effects. Finally we will compare treatment effects with other clinics.
Patients of the HF clinics were asked to complete several self-assessment questionnaires and surveys at their arrival (before treatment) and at their release (after treatment) within a few days. The reception survey also asked for the sociodemographic situation. For a followup, some of the patients received a selection of questionnaires again one year after treatment (see Figure
The data pool describing a patient in the therapeutic setting of Heiligenfeld consists of the patient’s history and demography, the psychological and somatic diagnostic data provided by the therapist, and a number of questionnaires that were distributed to the patients. Questionnaires were completed by the patients before, directly after, and one year after treatment.
Patients entering the clinic were initially diagnosed by a medical doctor as well as by a psychologist in accordance with ICD-10. Based on the main psychological diagnosis, they were assigned to one of 13 treatment paths (for details see below). A treatment path is defined by the compilation of therapeutic interventions and groups depending on the main diagnosis for a more specific treatment of various diagnoses.
Our analysis is based on several self-assessment questionnaires, which are used to measure symptoms, work-related coping behavior, personal beliefs, and habits. The questionnaires were in German, the native language of the participants, and are presented below.
The HF clinic is part of a benchmarking system, developed and managed by the Institute of Quality Development in Psychotherapy and Psychosomatics (IQP, Munich, Germany). The system is called BaDo (Psychosomatic Basis Documentation) [
The above-presented set of variables and their organization allowed for testing the following specific research questions. How are the clinic-specific salutogenetic concept and the underlying values realized in the Heiligenfeld clinics? How does the concordance between the ethical concept and its perception by the patient influence the patient’s treatment success and salutogenesis? What treatment effect sizes were achieved? How are treatment effects dependent on the diagnosis or the treatment path? How sustainable are treatment effects? How do treatment effects compare to other clinics?
Table
The data sets and its underlying patient population are listed for each research question.
Research question | Inventory/data set | Subgroups | Number of patients |
---|---|---|---|
Q1 | HFV total and 14 items | All patients* | 7987 to 8170 |
| |||
Q2 | HFV total (i.v.) |
All patients* | 7259 |
| |||
Q3 | ISR (pre + post) (total + subscales) |
All patients* | Up to 8068 |
| |||
Q4 | ISR total effect size |
Patients in 13 paths* | 38 to 6962 |
| |||
Q5 | ISR pre + follow up |
Follow-up return* | Up to 1046 |
| |||
Q6 | ISR (pre + post) (total + subscales) |
Patients in 2011 | 1181 (HF) |
*As many patients as possible were included who completed the particular questionnaire scale partly or fully. i.v.: independent variable; d.v.: dependent variable.
Raw data were received from HF. Data processing and the calculation of questionnaire factors and scales were performed according to the official manuals. Missing item data were replaced by the average of item values which contribute to a specific factor. This means that for a valid factor at least one item has to be answered, otherwise the patient is excluded from the data pool contributing to a specific factor. The calculation of effect sizes requires a normal distribution. Unfortunately, Lilliefors tests as well as Kolmogorov-Smirnov tests applied to the data of the factors of the ISR, TPV, VEV-K, and GV failed. However, a comparison between normal and actual distribution resulted in an acceptable fit (10% error) in the probability range between 10 and 90% for the ISR, TPV, and GV. The VEV-K showed an acceptable fit in the range between 15 and 80%, and the HFV only provided a good fit in the probability range of 25 to 75%, that is, values between 3.5 and 3.9 for the HFV total factor. The standard error bars in the VEV-K scores therefore have to be treated with caution. Cohen’s d effect sizes of ISR factors and the TPV were calculated using the difference between pre- and posttreatment means (or follow-up mean) and divided by the pooled standard deviation.
Our study focuses on the inpatients in three clinics of the Heiligenfeld group which are all located in Bad Kissingen/Germany. Two clinical categories can be distinguished: one clinic for patients with a public health insurance (Fachklinik, 37.0%), and two clinics for patients with private health insurance (Parkklinik, 52.5% and Rosengartenklinik, 10.5%).
After exclusion of patients who stayed less than 14 days in the hospital data from over 11,200 patients treated during 1/2007 to 1/2013 were available. This set consisted of 69% female and 31% male patients, aged between 17 and 74 years (average age 45.9 ± 9.9 years) who stayed in the hospital for 54.7 (±19.5) days on average. 57% of them had a higher educational level (university or similar). Patients with various psychosomatic disorders were included. The follow-up survey one year after treatment was effectively completed by about 1050 patients depending on the questionnaire.
The most frequent treatment paths were depression (86.3%,
The distribution of patients into the various treatment paths is illustrated. Several diagnoses per individual are possible.
First, the patients’ ratings are reported indicating how well they find the HF values realized in the clinic before in a second step (Q2) we explored how the concordance of patients with these values stands in connection with treatment success. Table
The mean rating is the average of between 7987 and 8170 patients on a scale between 1not applies and 4fully applies. SD is the corresponding standard deviation. The ratio evaluation columns show the percentage of patients who voted with “fully applies” (left) or with “partly or fully applies” (right).
Heiligenfeld values | Mean rating | Ratings in % of patients | |||
---|---|---|---|---|---|
Fully applies4 | Partly applies3 | Rather not applies2 | Applies not1 | ||
Heiligenfeld is a… | |||||
human place | 3.84 | 84.6 | 14.5 | 0.6 | 0.3 |
healthy place | 3.63 | 65.9 | 31.1 | 2.6 | 0.4 |
place of healing | 3.66 | 68.5 | 28.7 | 2.5 | 0.3 |
place of love | 3.43 | 50.4 | 42.7 | 5.9 | 0.9 |
place of wholeness | 3.48 | 55.9 | 37.0 | 6.1 | 0.9 |
place of spiritual growth | 3.42 | 54.0 | 35.9 | 7.8 | 2.3 |
place of community | 3.78 | 79.1 | 19.1 | 1.5 | 0.3 |
place of mindfulness | 3.49 | 54.4 | 40.4 | 4.8 | 0.5 |
I think, in Heiligenfeld… | |||||
fundamental rights are respected | 3.77 | 78.0 | 20.6 | 1.2 | 0.2 |
people are being acknowledged and appreciated | 3.74 | 76.4 | 21.3 | 1.9 | 0.3 |
the essential of life is touched | 3.68 | 71.1 | 25.7 | 2.9 | 0.3 |
responsibility is experienced | 3.54 | 59.4 | 35.9 | 4.2 | 0.5 |
life is respected | 3.81 | 81.8 | 16.9 | 1.2 | 0.1 |
empathy is experienced | 3.73 | 75.7 | 22.1 | 2.0 | 0.2 |
| |||||
Mean of all items | 3.64 | 91.3 | 8.3 | 0.3 | 0.0 |
The question how the perception and concordance of the HF values connect the therapeutic effect was answered by a regression analysis with the HF value ratings as independent variable and (a) the total score of the ISR change before and after treatment or (b) the total VEV-K score after treatment as dependent variables. The results are visualized in Figure
The distribution of the total HF value ratings is visualized in a scatter plot which on the vertical axis represents either the ISR total change (a) or the VEV-K total change (b) after versus before treatment. A linear fit function is inserted to illustrate the dependences of HF value rating and therapeutic effect.
The VEV-K total score showed an even more prominent dependence with the HF value ratings by offering an
As a result of this analysis we are supposing that the HF values might have a salutogenetic effect in the sense that physical and mental health improvement as well as symptom reduction is connected to the subjective experience of the realization of values. Due to the unknown causation and more generally speaking, we can only conclude that those patients who rated themselves as having achieved greater improvements could find the HF values more present in their own personal experience of the environment.
Answering the question “who are the patients who cannot find those values realized in the clinical setting?” already before treatment requires a closer look to the pretreatment survey completed at reception. Here, contentedness ratings as well as the supplementary variables of the ISR significantly differed between “agreeing” and “disagreeing” patients to the realization of the values (
The treatment effect of a salutogenetic concept can be measured through symptomatic assessments but should also include measures of general health parameters. The improvements of the latter are given by the VEV-K and the GV questionnaire data. The ISR differences between pre- and posttreatment data depict the symptom reductions during treatment. For each factor of the ISR, only those patients were included in the analysis who completed the respective items and who showed a symptomatic load already at reception on this factor. This analysis included patients of all diagnoses. The number of patients included in each factor is listed on the right in each figure. Table
Symptom-related treatment effects. Effect sizes (Cohen’s
ISR total ISR subfactors TPV total | Pre-prost-treatment | |
---|---|---|
Effect size (Cohen’s |
|
|
Total score | 1.06 | 8068 |
Depression | 1.37 | 7732 |
Anxiety | 0.82 | 6263 |
Obsession | 0.82 | 4995 |
Somatic symptoms | 0.91 | 3720 |
Eating disorders | 0.71 | 3335 |
All supplementary factors | 0.95 | 6549 |
Suicide | 0.90 | 2070 |
Problems with sleep | 0.69 | 7144 |
Appetite | 0.88 | 3083 |
Memory | 0.73 | 7034 |
Negative memories | 0.67 | 4372 |
Psychological problems | 0.83 | 5450 |
Alienation | 1.08 | 2862 |
Sexuality | 0.70 | 4529 |
Traumatic experiences | 0.74 | 1834 |
TPV total | 0.69 | 7215 |
Symptomatic changes after treatment were assessed using the ISR self-rating questionnaire. Here the scale differences were used as a measure with standard error bars.
The health change questionnaires GV and VEV-K present a quite balanced picture in post-treatment changes across the various factors as can be seen in Figure
Changes of general health parameters after treatment. Population sizes are indicated on the right vertical axis. The horizontal axis indicates the changes in points on the scale.
The question arises how such a treatment concept can deal with the variety of different diagnoses. This question was answered by looking at the symptom and health-related changes of patients’ self-ratings who participated in different treatment paths. In Figure
Total symptom rating for each of the 13 treatment paths. Effect sizes (Cohen’s
Diagnosis-specific treatment path | ISR total scores | |||||
---|---|---|---|---|---|---|
Post-pre-treatment | Follow-up pretreatment | |||||
Effect size |
|
Effect size |
|
Effect size |
|
|
Depression | 0.91 | 6962 | 1.07 | 677 | 0.87 | 677 |
Anxiety disorder | 0.98 | 1649 | 1.13 | 178 | 1.00 | 178 |
PTSD | 0.64 | 1156 | 0.82 | 109 | 0.76 | 109 |
Eating disorders | 0.95 | 726 | 0.93 | 76 | 0.70 | 76 |
Somatoform pain disorders | 0.76 | 622 | 0.77 | 69 | 0.63 | 69 |
Borderline | 0.71 | 529 | 0.81 | 71 | 0.69 | 71 |
Addiction | 0.97 | 480 | 0.83 | 43 | 0.96 | 43 |
Obsessive-compulsive dis. | 0.95 | 174 | 0.78 | 23 | 0.85 | 23 |
Narcissistic personality dis. | 0.60 | 165 | 0.81 | 29 | 0.56 | 29 |
Psychosis | 0.80 | 131 | 1.17 | 13 | 0.60 | 13 |
Suicidal thoughts | 0.86 | 120 | 1.62 | 9 | 0.82 | 9 |
Religious/spiritual crisis | 0.82 | 92 | 1.13 | 12 | 0.91 | 12 |
Self-harming | 0.92 | 38 | 0.51 | 2 | 0.24 | 2 |
(a) The VEV-K total score changes depending on the treatment path which is related to the main diagnosis. (b) The effect sizes in the improvement of ISR total scores of pre- versus post-treatment in dependence of the treatment path which is related to the main diagnosis. In red, the corresponding standard error bars are depicted.
The remaining groups PTSD, somatoform pain disorders, borderline, and narcissistic personality disorders revealed medium effect sizes
While the term “healing” is often perceived as an act of repairing a dysfunctional human system, the term “salutogenesis” includes the understanding of empowering the patient to generate a state of health. A salutogenetic treatment should additionally provide tools and abilities to a patient to retain and generate health. In this perspective a salutogenetic treatment should present a basis for long-term stability in health-related parameters. In HF, a follow-up survey has been sent out to some patients one year after treatment. Figure
Comparison of changes in general health parameters directly after treatment and in a one-year follow-up survey. Between 950 and 1046 patients are averaged in each bar.
The connection between symptom reduction and the improvement in health-related factors of a wider concept of health as assessed by the VEV-K and GV was exemplarily studied in patients who participated in the depression path. Changes on the depression scale of the ISR were correlated with the factors of the VEV-K and GV using Spearman’s rank correlations. As shown in Table
Correlational analysis between depression symptom changes of patients with depression as main diagnosis and health change factors of the VEV-K and GV. Post-pre-treatment:
Health change factor |
Spearman’s |
Spearman’s |
---|---|---|
VEV_Total | −0.39 | −0.36 |
VEV_Coping/future | −0.38 | −0.37 |
VEV_Social security | −0.33 | −0.28 |
VEV_Resting/satisfaction | −0.35 | −0.31 |
GV_Total |
|
|
GV_Somatic | −0.33 | −0.39 |
GV_Mental |
|
|
GV_Self-esteem | −0.33 |
|
GV_Coping |
|
|
Post-pre-treatment:
Bold values indicate correlations larger than 0.4.
The above-reported results can be set in relation to the treatment effects of 14 other German clinics.
Table
Psychosomatic treatment effects in comparison between HF and 14 other German clinics. A paired
Scale | HF clinic mean (SD) | Other 14 clinics mean (SD) | Significance test |
---|---|---|---|
ISR total pre-post | 0.58 (±0.48) | 0.52 (±0.51) |
|
ISR depression pre-post | 1.35 (±0.97) | 1.18 (±0.96) |
|
VEV-K | 129.3 (±26.9) | 128.3 (±26.7) |
|
GV physical | 4.42 (±0.73) | 4.22 (±0.77) |
|
GV psychological | 4.69 (±0.87) | 4.44 (±1) |
|
GV self-esteem | 4.68 (±0.83) | 4.38 (±0.92) |
|
GV social problems | 4.07 (±0.88) | 4.05 (±0.94) |
|
GV private relationships | 4.4 (±0.83) | 4.26 (±0.95) |
|
GV occupational | 3.89 (±0.83) | 3.77 (±0.92) |
|
GV motivation | 4.39 (±0.79) | 4.32 (±0.88) |
|
GV comprehension | 4.88 (±0.84) | 5.2 (±1.13) |
|
GV future orientation | 4.51 (±0.9) | 4.34 (±1.02) |
|
GV well-being | 4.68 (±0.85) | 4.4 (±1.02) |
|
GV daily life requirements | 4.33 (±0.81) | 4.19 (±0.88) |
|
Starting with the assumption that a sustainable form of healing in a psychosomatic inpatient clinic requires more than the focus on symptom reduction, we presented one approach of a psychosomatic hospital in Germany that follows a holistic and specific salutogenetic approach. It should be noted that several German psychosomatic clinics follow similar salutogenetic ideas to a certain extend. The approach of the HF clinics was formulated in a guideline that includes a set of values intended to generate salutogenetic behavior and experience within the whole clinic environment. Also, patients are supported in living a health-oriented and self-directed life.
In the first part of this study our aim is to demonstrate whether and how several aspects of human values underlying the holistic concept have been implemented in a clinical setting for inpatient treatment of psychosomatic disorders. Therefore, the clinic has developed a self-assessment questionnaire. The results suggest that salutogenetic values have not only been formulated in the clinical guidelines but were also successfully transferred to the therapeutic environment. This self-assessment shows that 90 to 99% of the patients agreed to the statement that those values were realized and present in the clinic. Unfortunately, the high concordance with the questionnaire statements moved the distribution towards the upper end of the scale and created a saturation effect with a non-Gaussian distribution. Therefore, we did not perform correlational analysis with treatment effects. Nevertheless, the high number of patients allowed a linear model to be fitted to investigate the connection between patients’ perception of the clinical value codex and the therapeutic effect. The strong dependency between symptom reduction and value ratings and the total score of health changes (VEV-K) suggests that it is important for a hospital to offer a supportive environment that allows patients to get involved with salutogenetic factors. Those factors were intended to be realized not only within therapeutic sessions but also in other areas of the patients’ daily life in the clinic. The statements about concordance with the ethical values may be biased, due to patients wanting to express their gratitude towards the clinic. We took a closer look at those patients with low concordance ratings, and found that they fit a certain profile. They were less content and offered a higher load in symptoms, however, not in depression. From 56 “disagreeing” patients, 89% had sleeping problems, 88% memory problems, 70% had problems with their sexuality, 36% were suicidal, and 32% suffered from PTSD symptoms.
Data acquisition for measuring the therapeutic outcome was performed with self-assessment inventories as well. We are aware of the short comings of self-assessment questionnaires but want to point out that, particularly in the context of psychosomatic disorders, perceived improvements in health are very valuable to the individual, regardless of the explanations for these changes such as possible placebo effects or insufficiencies of the questionnaires. It should be noted that an F diagnosis assessed by a therapist also is based on patients’ reports. Therefore, it is not a surprise that, according to the IQP, the self-assessments are roughly in line with the diagnoses given by the therapist. For some measures, especially those asking for mental states and subjective experiences, one has to rely on the self-ratings for now. Nonetheless, the analyses would benefit from an inclusion of more popular questionnaires such as the SCL90. However, the clinic intends to use questionnaires with fewer items and prefers license-free inventories. Further, it might be interesting to additionally measure the HF values in other clinics to find out whether the ceiling effects occur there as well and for a better understanding and possible confirmation of the importance of these factors.
We have introduced salutogenesis as a complex construct with many different aspects. Salutogenetic factors are inherent in concepts such as mindfulness, spirituality, social support, and the sense of coherence that comprises comprehensibility, manageability, and meaningfulness. Aspects of those concepts were found to a certain extend in the HF guidelines and/or the treatment concept of the clinic. However, despite not all of the salutogenetic factors were explicitly assessed in the questionnaires, the questionnaires VEV-K, GV, and the TPV used in HF during several years cover those aspects to a certain extent.
The sustainability of improvements in a 1-year follow-up was about 80% and can be regarded as high. However, these data are only available for a small proportion of total patients. Comparing the posttreatment data of the follow-up group with the posttreatment group of all patients, those patients who returned the follow-up questionnaire already showed higher treatment effect sizes at release time. Therefore, it is questionable whether patients with lower treatment success—often those with more severe symptoms—would also display such high sustainability. The fact that treatment paths involving patients with more severe symptoms showed a far lower sustainability argues against it.
In direct comparison between HF and the other clinics it should be noted that while the distribution in age and gender was comparable, the number of patients with a private health insurance was 70% in HF but less than 10% in other clinics. This also links to a significantly higher education level in HF compared to other clinics. Furthermore, the average treatment duration was about 14 days longer in HF than in other clinics (57 days versus 43 days). Since both parameters, educational level and treatment time, are positively related to health improvement, the superiority of the HF concept over the concept of other clinics cannot be definitely stated and would require further analysis. Considering all results, the data suggest that the perceived successful implementation of a guideline that is based on the idea of salutogenesis affects patients’ health state in a positive way. However, as other clinics also follow more or less salutogenetic approaches, the comparison between clinics is not very meaningful.
Successful implementation of a salutogenetic concept requires, through the entire hospital staff, the promotion of an autonomous development and dialog of values, including meaningfulness and spirituality. Patients might then profit from the inner attitude they perceive from everybody within the clinical environment, possibly enabling them to integrate this experience as an ideal for their own coping strategies leading to an increase of their SOC. An interesting future study could be to explore the effect of the inner attitude of the therapist on the treatment success of their patients.
The highest symptom reductions were achieved in depression, which was also most commonly diagnosed and also was the most frequent treatment path. A correlation analysis of patients with depression revealed medium but sustained correlations between the improvements in coping-related aspects and depressive symptoms. This supports the idea that proactive behavior (manageability in the SOC model) is a key feature for health.
The TPV measures different facets of religiosity related to a transpersonal trust. The pre-post-treatment comparison only shows medium effect sizes. Probably, the gain in transpersonal trust seems not to be as strongly promoted in the clinic as other forms of spirituality. This assumption can be derived from the observation that 90% of the patients agree that HF is partly or fully a place of spiritual growth. To address this issue, additional tests investigating meaning and spirituality, for example, LEBE by Schnell and Becker [
With these findings we would like to propose the idea that a salutogenetic treatment requires a value driven social environment that provides a learning atmosphere for the acquisition of health supporting skills, behavior, and attitudes and finally leads the body and mind into a state of self-recovery and to the development of skills for further “health development” in order to continue the recovery process in daily life. We therefore propose that salutogenesis should be initiated in the clinical environment and in the therapeutic processes, with a continuation beyond the clinical stay. This goal can only be met when patients learn in the clinic setting how to live a health promoting, self-directed, and fulfilled life.
The authors thank the Heiligenfeld Kliniken, Bad Kissingen, Germany, for providing data and further information. The authors also are thankful to the IQP Munich for being able to report comparative results.