Patient satisfaction with primary care: an observational study comparing anthroposophic and conventional care

The inclusion criteria for physicians in the AM group were working as primary care provider for at least two days a week and membership in the Swiss Medical Association for Anthroposophic Medicine (VAOAS), which has the following prerequisites: Completed specialist training in a CON discipline, 360 hours of training in AM (as an assistant anthroposophic doctor in a clinic, practice, hospital department or independently together with a mentor), and participation in a study group of physicians for AM. Moreover we only included physicians, from whom we could sample at least five patients.

51 of the 134 members of the VAOAS, who were invited by letter, participated in the study. 32 met the inclusion-criteria of working as primary care provider for at least two days a week. In PEK II we matched 71 CON physicians who were not listed as members in any CAM medical association in Switzerland to AM physicians using a stratification technique based on geographic distribution. Three questionnaires evaluated structure, process and outcome of care.

Patients were classified according to the method of treatment they chose into the AM and CON groups. We only included patients over 16 who gave their written consent. The ethics committee of the Canton Berne raised no objection to the study. The study was conducted in compliance with the Helsinki Convention.

The structural data on the physicians and their practice were taken from PEK I. We developed the questionnaire in German, French and Italian together with an expert group of Swiss primary care providers specialized in CON and/or CAM.

Data collection took place in October 2002 and January, May and August 2003 on four different predetermined weekdays. Practice staff handed out a written questionnaire to all eligible patients consecutively visiting their practice on such days. Patients filled out the questionnaire in the waiting room prior to the consultation and returned it to the practice staff such that physicians were not aware of the content. The participating physicians were reimbursed with CHF 500 each.

Four weeks after, patients were sent a second questionnaire directed to the perceived effectiveness of, and their satisfaction with, the treatment, fulfilment of their expectations, and whether they experienced adverse or positive side effects or other effects as a result of the treatment. They were also sent the Europep instrument [3]. Europep evaluates medical care with 23 questions and a five-point answer scale ranging from poor to excellent. Six Europep questions addresses "doctor-patient relationship and communication", five questions addresses "medical-technical care", four questions addresses "information and support to patients", two questions addresses "continuity and cooperation", and six questions addresses "facilities, availability and accessibility".

All data were recorded using a relational database. Forms filled out by patients and physicians during consultations were coded and recorded manually. The questionnaires were machine-readable and were scanned by the Swiss Federal Office of Information Technology using Optical Character Recognition (OCR).

Data derived from the Europep questionnaire were reduced to a two-level scale with the most favourable answer category coded as one and all other non-missing categories as zero. These data were analyzed using hierarchical multivariate procedures for each individual question [31]. In addition to the AM group, patient age and gender were included in the models as additional factors. Similar models were used to evaluate the probabilities of complete symptom resolution, complete fulfilment of expectations and of being very satisfied with the treatment. All analytical procedures accounted for non-independence of observations at the practice level and 95% confidence intervals (95% CI) of means proportions and odds ratios were calculated accordingly.

The level of significance was set at p < 0.05 throughout the study and SAS 9.1 (SAS Institute Inc., Cary, NC, USA) was used for all calculations.

The 71 CON and 32 AM physicians (see Table 1) did not differ significantly in age and clinical experience, but compared with CON more AM physicians were German speaking, female, worked part-time, in group practices, and in inner cities. Nearly all CON practices had a laboratory, ECG, X-ray and ultrasound, whereas most AM practices were only equipped with ultrasound. The consultations of AM physicians lasted on average seven minutes longer than those of CON physicians.

Table 2 shows socio-demographic data, the self-rated health status of the participating patients, their reasons for consultation and their expectations. AM patients were primarily German speaking, female and better educated and more frequently reported chronic health problems than CON patients. Significantly more CON patients chose their physician for pragmatic reasons (for example, geographic proximity of the practice), whereas AM patients were more likely to choose their GPs based on the preferred procedure. The self-assessment of the patients of their illness in both groups was similar; however, AM patients had on average a higher risk of mortality, as measured by the Charlson index. Despite the higher risk of mortality, AM patients more frequently expressed the expectation of being healed.

The diagnosis of the patients in the two groups is shown in Table 3. There was a significant difference in the distribution of diagnoses between the two groups. AM patients were diagnosed more often with neoplastic diseases (ICD10 Codes C00-D48), whereas CON patients were twice as likely to have diseases of the circulatory system, injuries, poisoning and endocrine and metabolic diseases. With respect of the distribution of co-morbidity, there was no statistically significant difference between the groups (p = 0.398). Slightly more AM patients (65.01%) had two or more diagnosis as compared to 60.67% for the CON group. AM patients had significantly (p < 0.000) higher scores in the Charlson co-morbidity index [32], which indicates that they had higher mortality risks.

1946 patients of 103 AM and CON GPs were evaluated, representing a proportion of returned questionnaires of 45.8% of the 4249 patients. 51.2%, of the AM patients responded as compared to 43.8% of the CON patients. Altogether, more females (49.8%) than males (40.9%) and more chronically ill (50.5%) than non-chronically ill patients (42.2%), responded to the survey. Responders were on average 53.3 years old, non-responders 9 years younger.

As shown in Table 4, 56.1% of patients receiving AM treatment from their GP were significantly more satisfied with the overall treatment as compared to 43.4% in the CON group. 38.7% of the AM patients reported that the treatment completely fulfilled their expectations (vs. 32.6% for the GPs using CON). AM patients reported significantly fewer adverse side effects (9.3% for AM v.15.4% for CON). In 31.7% (vs. 17.1% in the group treated with CON) patients noted other positive effects and patients receiving AM treatment only complained of other negative effects in 3.0% of the responses (vs. 6.8% for the patients of GPs employing CON).

The characteristics of better satisfaction and higher likelihood of successful treatment as well as the absence of negative side effects were independent of age and gender of the patients.

Table 5 sets forth the percentage of the patients who gave the highest rating ("excellent") in the Europep instrument 4 weeks after their visit. AM patients valued their relationship and communication with their physicians more than did CON patients. As to the factors whether physicians make them feel they had time during the consultation (76.5% vs. 61.7%, p < 0.001), physicians' interest in the personal situation of the patients (74.6% vs. 60.3%, p < 0.001) and that the physician was listening to them (80.0% vs. 67.1%, p < 0.001), differences between the AM and CON group were highly significant. AM patients evaluated significantly more often that their physician made it easy for them to tell him or her about their problem (71.6% vs. 62.9%, p = 0.023) and that the physician involved them in decisions about their medical care (67.8% vs. 58.4%, p = 0.023).

In addition, more AM patients than CON patients ranked their physicians "excellent" concerning the giving of information and support, helping them to deal with emotional problems related to their health status (61.3% vs. 49.7%, p = 0.004), telling them about what they wanted to know about their symptoms and/or illness (69.9% vs. 60.2%, p = 0.005) and explaining the purpose of tests and treatments (68.0% vs. 60.2%, p = 0.044).

A much higher percentage of the AM patients valued the thoroughness of the GP (70.4% vs. 56.5%, p > 0.001). The patients receiving CON treatment reported that their GPs more frequently provided preventive services, such as screenings, health checks and immunizations (48.7% vs. 41.5%).

As in studies investigating CAM [4, 33], AM in other countries [12, 16, 35] and in Switzerland [36, 37], urban, middle-aged women (30 to 50 years) with higher education were overrepresented in our AM group (see table 2). Highly educated patients may be better able to follow the AM approach, actively taking part in their treatment. They also might adapt better to stress and changes brought about by the illness, for example through a meaningful support or a positive interpretation of their diseases [14, 15]. As AM patients have shown to be more convinced that their lifestyle has an impact on their health [35], these patients with a more active approach in managing their problems may have a greater sense that their condition is manageable and this increases satisfaction [14, 18]. AM therapy does not work without the cooperation of patients. Therefore, some AM physicians only accept patients who are highly motivated, responsible and "psychologically mature" enough to work with AM [12]. This inherent selection could explain some differences in the patient groups.

The AM patients in our study, as in prior studies [12, 21], suffered more frequently from chronic diseases of the musculoskeletal and respiratory system, mental and behavioural disorders and cancer than CON patients, who suffered more often from diseases of the circulatory system (see table 3).

For these chronic illnesses of our AM patients, as well as for "non life-threatening" diseases, such as psychosomatic or functional/psycho-vegetative disorders or certain pain-syndromes (e.g. migraine) with feelings of ill health, or with marked subjective symptoms for which no severe organic disease is present, further CON diagnostics and treatment were unsatisfactory because of ineffectiveness, adverse effects, or non-compliance [38], or were not indicated.

As chronic illness is the most common cause of disease burden worldwide (often associated with co-morbidity) [4], successful AM treatment could result in a reduction of health care costs [39], in particular, since CAM can lead to improvement or bring relief in the areas of clinical practice in which CON treatment is not fully effective (e.g. musculoskeletal problems, chronic pain, eczema, depression, cancer, etc.) [37]. In patients with these types of common illnesses, CAM methods are often more beneficial, although the cost-effectiveness is disputed [39, 40].

A common definition of patient satisfaction is "fulfilment of treatment expectations." Patients choose AM for its holistic and person-centred approach that is tailored to individual needs, or in situations of limited effectiveness of CON in case of chronic diseases and cancer [17, 23]. They expect the facilitation of personal learning and development [12], wish to be involved in the management of their illness [41], or want to do everything possible to fight an incurable disease [42].

Further aspects related to the specific AM approach, such as the quality of physician-patient relationship, the use of natural treatments and remedies with few side effects, the activation of self-healing through art therapies, and the wish for the holistic AM therapy [17, 23] seem to be key reasons that patients seek AM therapy.

Perhaps some of our AM patients also belong to these "expert patients", who exchanges for the public health system invisible in networks, self helping groups or chartrooms wishing to be involved in the management of their diseases [41].

A growing number of patients, reject the traditional authoritarian and pathologically oriented role of western CON physicians, feel misunderstood, incompletely advised or treated unsatisfactorily. These patients tend to change to CAM methods, which were closely linked to their salutogenitic needs and their expectations to be equal partners with the physicians in treatment decisions [41].

The higher expectation of healing as opposed to relief of symptoms that we found in the AM group (see table 2) may be related to the AM theory that illness is an imbalance among the forces of body, mind and spirit, which can generally be rebalanced or even healed [11]. This may give patients a degree of optimism [14].

A further positive factor for AM may be significantly fewer adverse side effects. While CON drugs are specifically prescribed for particular physical pathologies and have strong effects and side effects, AM treatments aim to activate the whole person, restore inner balances and activate self-healing capacities at different functional levels [10, 12]. This is accomplished by therapies to which the whole person reacts with body, soul and spirit, such as music- or art therapy, eurythmy, or massage. Also, in accordance with its principles of "salutogenesis" [18] and "hygiogenesis" [10, 43], AM attempts to specifically activate "life forces", which are considered responsible for all processes of growth, vitality, self-healing, self-regulation, adaptation and regeneration [10]. This is done, as in homeopathy or herbal medicine, through special pharmaceutical preparations from minerals, plants or animal substances (e.g. potentization), aimed at eliciting specific effects. Art therapies and „mild“ agents were known to have only few side effects [10, 20] and as such could have contributed to the higher patient satisfaction in the AM group [16, 44, 45]. This is in line with the observational evidence of high safety and sustainable effects of the treatment with AM on perceived symptoms and to improve quality of life in chronic diseases, including advanced cancer and depression [20].

Further factors that may lead to higher patient satisfaction are the patients' positive attitude towards AM and its theories as well as their expectation or „belief“ of likely benefit. This can be seen as a placebo-response, but underlying this there may also be significant optimism [46] and trust [12, 47] of patients who had good experiences with AM or had heard about others who did so, especially in those diseases where CON treatments were at their limits [38]. The fact that AM physicians have the option of prescribing both conventional and anthroposophic therapies might also strengthen trust in AM treatment.

"Other positive effects" in the AM group were perhaps personal experiences with the therapy or factors associated with becoming proactive in their own treatment. Patients may have described a "build up effect" or a "feel good factor" after AM appointments in that patients expressed feeling more positive when they came out than when they went in [12]. This may reflect AM therapy meeting the expected health needs of our patients through a greater focus on individual responsibility and providing deeper-level explanations of health and illness, linking psychological and physical dimensions, which may help to cope with the illness, finding a new meaning of life or self-development [14].

That CON patients mentioned "other negative effects" and "adverse side effects" more often could reflect a higher risk of side effects or drug interactions with conventional drugs or with drugs taken without knowledge of the physician. Further aspects could be the missing consultation time or that many of the patients' real problems could not be solved by a non-holistic approach.

To our knowledge, our study is the first to investigate patient evaluation of their primary care providers comparing AM to CON, using the Europep instrument to provide a subjective assessment of different aspects of care provision in positive and negative terms. The Europep instrument queries judgments by patients, in contrast to satisfaction, which assumed to be a (general) emotional reaction to a specific situation [3]. In international comparisons of Europep results, Swiss patients are known to give high rating scores to their physicians (often the best or second to the best ranking) in items of the dimensions that we classified as "relationship and communication" [3], but these absolute higher ratings in Switzerland would be expected to affect both groups equally and not to bias the comparison of AM and CON in this study.

The structure of our AM practices, which were predominantly situated in the German speaking part of Switzerland, was similar to the structure of other CAM practices [36] and AM practices in other countries [12]. Namely, these CAM and AM practices were more frequently group practices, with more part-time physicians and with less technical equipment than CON practices (table 1). In addition, the CAM and AM practices offered more patient-centred and individualized treatment modalities [40]. The central location of the practices could be explained by the need to serve geographically dispersed patients. As expected, our AM physicians were more likely to speak German.

There has much been written about the setting in which the clinical encounter between a patient and a healthcare professional takes place, which is seen as the core activity of medical care [28, 48] and how the physician can contribute to good communication [26]. In the practices of our study, these effective communication and affective relationship dynamics were generally known and certainly implemented, which contributed to the high ranking for both groups. Good communication is particularly important for chronically ill patients, since it improves patient compliance and thus improves the quality of care [49].

Physicians practicing AM have longer consultations, taking an extended history, addressing constitutional, psychosocial, and biographic aspects of patients' illnesses, and selecting optimal therapy [12]. The consultations lasting seven minutes longer in our AM group seem short when considering the goals and methods of AM as an extension of CON [12]. In that respect and in light of the characteristics of their patients, AM physicians seem to work efficiently, since even CON physicians tend to have longer consultations with chronically ill patients [50].

In our study, AM physicians showed higher interest in the personal situation, listened to patients more and involved patients more in decision making than CON physicians.

Our results were supported by results of another study, in which AM patients described their physicians to be good listeners with a 'calm' and 'unrushed' attitude and with a high degree of personal encouragement and interest in their patients [12]. In that study, patients who had employed both AM and CON treatment described the consultation with AM physicians more as a 'dialogue' or 'two way process' compared to the rush consultation with a CON physician [12]. Moreover, our results were consistent with the interpretation that the AM therapeutic concept leads to a more relational and supportive communication style [40].

That AM patients in our study more often felt involved in decisions about their medical care than CON patients could have resulted from a key aspect of AM therapy, which is to motivate patients to actively engage in their treatment and to take responsibility for addressing their health problems [12, 27]. It is known that shared decision-making is a challenge for both, patients and physicians: patients have to take more responsibility of their treatment, even in case of non-success, and physicians have to respect the patients wishes, even if they decide against the physicians advice. Deciding together improves quality of care [51].

AM patients were more content with data protection than the CON group. This may have positively reinforced (or may simply reflect) their trust in AM physicians.

AM physicians in our study explained tests and treatments more often, discussed symptoms and illness more often and helped the patient more often to deal with emotional problems than did CON physicians.

Patients increasingly demand medical advice as well as medical information in a manner and in language that they can understand and increasingly expect that their own concepts of self-healing be incorporated in decisions concerning therapy [41].

Patients externally referred to AM services are particularly impressed with the depth of information covered in consultations [12]. Often it is necessary to inform patients about the approach in AM consultations. To an AM physician, there is no simple catalogue of instruction to treat each particular disease. Rather, AM theory calls for the physician to imagine for each patient "flexible working pictures" implementing the theory of an integrative view of simultaneous interactions of the different subsystems accounted for in the AM understanding of health and illness [9, 10]. These pictures intend to help to find the right individual therapy. For the most part, AM practitioners are seen as knowledgeable and flexible in their approach to diagnosis and treatment [12]. AM physicians may give information about the imbalance, which led to the illness and may motivate their patients to participate actively in their treatment.

Patients with chronic conditions were especially appreciative of the positive approach taken by the AM doctors and a sense of hope they gave that not all possibilities of improving their condition had been exhausted [12]. However, AM physicians tend to give realistic information about there not being any 'guarantees of cure' [12].

Chronic or severe illness is often associated with feelings of depression, sadness and anxiety, with philosophical questions of ill health or psychosocial problems accompanying the illness. In this context, emotional problems can become acute. It is helpful to be able to speak to a person of trust about these problems. Through the formulation of the problems in words, the patient can gain a greater objective distance from the problems with which she is concerned. Together with the apparently closer physician-patient relationship of AM patients and the more intensive discussion, these factors could have led the AM patients to reveal more intimate information than did the CON patients. This may have been particularly important, since a higher percentage of AM patients compared to our CON patients suffered from chronic illnesses that required more thorough consultations.

Patients expected not only good counselling with time enough to communicate their concerns, to be informed and included in decisions about their illness, or physicians' secrecy, but also services for preventing diseases [3]. In the last category, the CON physicians scored significantly better than AM physicians; but AM physicians were more often judged as being thorough (table 5).

The CON practices of our study seem to reflect actual mainstream medicine in Switzerland that offers highly quality technical medicine combined with a personal service, and they also appear to follow current best practices in offering preventive services, such as screenings, health checks and immunizations.

In general, the physical dimension of illness remains the focus of CON. In light of their superior technical equipment, CON practices can perform the necessary diagnostics promptly and are able to diagnose and treat quickly acute health problems, e.g. of the cardiovascular system. This may be one reason that our CON practices treated more patients with cardiovascular diseases than the AM practices.

AM patients rated their physicians as more thorough, although they had less technical equipment. This may be due to the longer consultation time with more detailed medical and biographical history-taking and more intensive relationship and communication factors of AM physicians who were experienced in both CON and AM [10, 12].

Previous studies showed that patient satisfaction is less related to the therapeutic outcome [52, 53], and more to certain aspects of the therapeutic alliance [46, 54, 55]. Such an alliance presupposes a supportive physician-patient working relationship, in which the physician is seen as helpful, reliable, and successful in achieving common goals [54]. Our results concerning better ratings in "relationship and communication", as well as in "information and support" for the AM physicians, also support this theory. The more time-consuming AM procedure might better fulfil the needs of more critical or expert AM patients, who wish to be informed and take part in medical decisions, especially where the CON treatment is unsatisfactory or at its limit. The AM physician's personality, her empathy, and her willingness to communicate could be decisive factors for the higher patient satisfaction with AM (or even for its effectiveness). AM consultation appears in itself to be a therapeutic intervention working independently or synergistically with the prescribed therapy or agent for this group of patients.