Quality circles in ambulatory care: state of development and future perspective in Germany

Transcription

1 International journal for Quality in Health Care 1998; Volume 10, Number I: pp Quality circles in ambulatory care: state of development and future perspective in Germany FERDINAND M. GERLACH 12, MARTIN BEYER 12 AND ANGELIKA ROMER 1 'Department of General Practice, Quality Improvement Research Unit, Hannover Medical School, D Hannover and 2 AQUA Institute for Applied Quality Improvement and Research in Health Care, Gottingen, Germany Abstract Objective. To survey the quantitative development of quality circles (peer review groups; QC) and their moderators in ambulatory care in Germany, to describe approaches to documentation and evaluation, to establish what types of facilities and support is available and to assess opinions on the future importance of QC. Design. Cross-sectional survey using a standardized questionnaire and supplementary telephone interviews. Setting. All 23 German regional Associations of Statutory Health Insurance Physicians (ASHIP) were surveyed. Results. The total number of QC in ambulatory care in Germany increased rapidly from 16 in 1993 to 1633 in June 1996, with about 17% (range %) of all practicing physicians (112158) currendy involved. Throughout Germany, 2403 moderators were trained in 168 training courses by the qualifying date. Follow-up meetings were held or being planned in 20 ASHIP, with approximately 39% (23 95%) of the moderators participating. Systematic documentation of QC work was undertaken or planned in all 23 ASHIP, and 10 ASHIP carried out comparative evaluation, with at least five others planning to start it. The ASHIP promoted the work of QC by providing organizational (22) or financial (20) support, materials (20) or mediation of resource persons (16). Eleven ASHIP received grants from drug companies. ASHIP rated the future importance of QC as increasing (18) or stable (four), but in no case as decreasing. Conclusions and recommendations. The quantitative growth of QC in Germany is encouraging, but the extent of support and evaluation appears insufficient. Increased methodological support and facilitation, follow-up meetings on a more regular basis, improved documentation and evaluation of individual QC, and problem-oriented evaluation of their impact on health care are essential for further successful development. Principles, problems and solutions discussed may be relevant for similar QI activities in other countries. Keywords: ambulatory care, Germany, peer review, quality circles, quality improvement Although lagging behind neighbouring countries (The Neth- rates, and numerous diagnostic and therapeutic interventions, erlands or UK) by several years, changes in the German Over physicians work in private practices, but only health care system have resulted in increased efforts to 41.3% of them as general practitioners (GP) % are implement quality assurance and quality improvement meas- specialists. Patients have direct access to a broad range of ures. In 1993 new federal legislation was introduced to specialists, and ambulatory care physicians in Germany do provide a legal framework for quality improvement (QI) at not have patient registration lists. The vast majority of patients all levels of the health care system [1]. For the first time, QI (90%) are insured by statutory health insurance and only was given high future priority. 10% are privately insured. Physicians in private practice who Particularly in ambulatory care, it was necessary to find care for patients insured by the statutory health insurance adequate strategies to implement initiatives and involve a funds (over 98% of all ambulatory care physicians) are reasonable number of care providers. German ambulatory compulsory members of one of 23 independent regional care is characterized by a fee-for-service system, high contact Associations of Statutory Health Insurance Physicians Address correspondence to Dr F. M. Geriach, MPH. Tel: ( 49) 51 I /2746. Fax: ( 49) 51 I fmgerlach@aol.com 35

2 F. M. Gerlach et al. Table I Quality circles in German ambulatory care In Germany the term 'quality circle (QC)' is frequently used for groups that are synonymously described as 'peer review groups' elsewhere. In the context of ambulatory care we follow the definition of Grol and Lawrence who described 'peer review' as: 'continuous, systematic and critical reflection on their own and others' performance by a number of colleagues, with the aim of achieving a continuous improvement in the quality of patient care. Peer review is then an ongoing process involving the defining of criteria, evaluating performance, and implementing change; it focuses on continuous change'. [2] The participation of physicians in QC is entirely voluntary and is not connected with financial benefits. Although supportive of this approach to QI, ASHIP do not pressure physicians to join QC and disclaim control of contents and objectives. QC moderators (group leaders) receive a 2 day training course, but are themselves full-time physicians without any official charge. The courses usually offer training on group-oriented skills and try to increase knowledge of basic principles and techniques of QI. There are QC with members from one specialty (e.g. general practitioners working on waiting times in their practices) as well as interdisciplinary groups (e.g. anesthesiologists, dermatologists and surgeons trying to improve outpatient vein therapy). Only a few QC have started to integrate members from other professions (e.g. practice staff, community nurses, psychologists). Objectives and methods of QC have been suggested in guidelines issued by the federal ASHIP [3], but are decided upon autonomously by the individual group. Suggestions for objectives include the documentation of individual practice and comparison with colleagues' practice, the exchange of experiences, quality review of practice performance, and application/adaptation of existing guidelines (which are currentiy not obligatory in Germany). Rigid audits of the procedures and performance of individual practice, as are performed elsewhere, e.g. in the UK, are unusual. (ASHIP), which act as their corporate representative especially in negotiations about remuneration. ASHIP are also responsible for the functioning of ambulatory medical care and have an obligation to promote QI. The existence of these structures in the German ambulatory care system, has encouraged both governmental and professional bodies in Germany to favour the development of QI by peer review groups/quality circles (QC; see Table 1). While a few pilot QC in ambulatory care started as early as 1989, new rules and guidelines issued by the Federal Association of Statutory Health Insurance Physicians (Kassenarztliche Bundesvereinigung) in 1993 [3] put strong emphasis on internal QI through voluntary QC and provided favourable conditions for their further development. The rapid increase in QC (15 QC in 1993 [4]; an estimated 400 QC in 1994 [5]; and 631 QC in 1995 [6]) nevertheless depended strongly on the initiative and motivation of individual physicians. A feeling of isolation is common among physicians working predominantly in solo practices (about 85%) and particularly encouraged the development of peer groups. Funding through the Ministry of Health in 1993 supported the development of an influential and widely circulated handbook on QC that included experiences from other countries [4]. Training courses for moderators in implementing and co-ordinating local QC, manuals, documentation materials, videos, symposia and a system of regular supervisory meetings for QC moderators were developed [7 10]. In co-operation with AQUA (Institute for Applied Quality Improvement and Research in Health Care, Gottingen), the University Department of General Practice in Hannover has established the only information system for continuous documentation and assessment of QC in ambulatory care. This enables ongoing description of the development of QC, although only in a minority of four of the 23 regional ASHIP [11]. Thus, despite increasing numbers of QC in Germany since 1993, little was known about the extent of this development throughout the whole country. A more extensive descriptive cross-sectional study was thus required as a first step to obtain a general overview of this development. Although the results of this study cannot lead to a direct assessment of the impact of QC activities on the quality of care, they summarize the development of this approach to QI in Germany. In addition, they provide insight into the extent of ASHIP involvement in the guidance and evaluation of QC development. The main aim of our study was to survey the quantitative development of QC in ambulatory care in all parts of Germany, to describe approaches to documentation and evaluation and to assess what type of facilitation and support is available to these groups. The following QC-related areas were covered: distribution of groups and participants; number of moderators and training courses; 36

3 Quality circles in ambulatory care documentation and evaluation of QC by regional ASHIP; facilitation and support by ASHIP, pharmaceutical companies and professional bodies; predictions of subsequent quantitative development and opinions on future importance of QC. Methods To obtain an overall picture of all regional activities in Germany, we contacted the quality improvement co-ordinators (persons responsible for QI) in all 23 regional ASHIP, and carried out a complete cross-sectional survey among these key persons using a standardized postal questionnaire. The questionnaires were posted in June 1996 and received back between July and September (with one exception received in November 1996). The qualifying date for all statements was fixed as 30 June The questionnaire (11 pages) consisted of 33 questions in six sections: number of QC and participants (five questions), moderators and training courses (seven), follow-up meetings (seven), documentation and evaluation (six), facilitation and support of QC (five) and future perspective (three). Internal consistency was used as an indication of the validity of the responses. In addition, answers from four ASHIP were compared with data from the AQUA documentation system established in these regions [11]. To gather additional information and to clarify unclear statements, supplementary telephone interviews were carried out in September 1996 with the responders in all 23 cases. This comprehensive approach ensured that our data collection was as complete as possible in all our areas of interest. Results Personal contact with all responsible QI co-ordinators enabled us to achieve a response rate of 100%; questionnaires from all 23 ASHIP were received and analysed. Only a few questions remained unanswered by some responders. In six cases, no information was given about the number of physicians participating in QC. Therefore, this was estimated using the average number of participants per QC based on data from all regions and the number of QC in the ASHIP concerned. Within the regions of Bremen, Saxony-Anhalt, Schleswig- Holstein and Westphalia-Iippe additional data from the continuous information system for documentation of practising QC [11] was available for validation of the ASHIP statements. The comparison of the two data sources in these four regions showed very good agreement (e.g. the reported number of QC diverged by a maximum of 5%). Establishment and development: number of QC and participants By June 1996, the total number of QC in ambulatory care had increased to About 17% (19116) of all practising physicians ( ) were involved in these groups, with an average of 11.7 (range ) participants per QC. Because it is possible that a small number of physicians participated in more than one QC, the latter number might, in fact, be slightly lower. The absolute and relative numbers of QC varied widely among regional ASHIP. While there were 252 QC with 3024 participating physicians in Northrhine, only six QC with a total of 60 participants were known to the ASHIP in Berlin (see Table 2). As the number of physicians registered in ASHIP ranged from 720 in Trier to about in Bavaria, it was necessary to look at the number of QC or participants in relation to the total number of physicians in each region. The size of all regional ASHIP and the exact number of participating physicians is shown in Table 2 (first and third columns). Participation of physicians ranged from 1.0% in Berlin to 52.1% in Trier. In general, the relative number of QC was lower in East Germany. On the qualifying date, there were 1633 QC and ambulatory care physicians in Germany, that is 14.6 QC per 1000 physicians, with a range of depending on the region. The ASHIP were asked to predict whether the number of QC would increase, decrease or remain stable. The majority (19 ASHIP) predicted an increase, and four expected the number of QC to remain stable. None of the ASHIP believed that the number of QC would decrease. While general practitioners pioneered the development of QC [8], physicians from the entire spectrum of specialities involved in ambulatory care have begun to participate in this form of QI. QI co-ordinators in the regional ASHIP estimated that roughly 68% (» = 1113 of 1633, range %) of the known QC met the requirements outlined in the guidelines issued by the National Association of Statutory Health Insurance Physicians (Table 1) [3]. Moderators and training courses Moderators are ambulatory care physicians with a special interest in QI who are willing to take on this function as honorary volunteers. By the qualifying date, 2690 moderators were registered with ASHIP, of whom 842 were not leading a QC at the time. Thus, 1848 moderators were active in 1633 QC. Only 189 physicians were acting as moderators or comoderators without having completed training. The remainder (2501) had been qualified by special training sessions or former accomplishment. The majority of moderators (2403) participated in one or more of the 168 training courses arranged by the ASHIP throughout the country. Over twothirds (114) of these courses were held during the 12 months preceding our survey. An additional 93 moderators were known to have been trained by other organizations, such as professional bodies. Because a number of different expert groups offered training courses for moderators in Germany [9,10], it remained unclear what underlying concepts were used by them and what implementation effects resulted. 37

4 F. M. Gerlach et al. Table 2 Number of physicians participating in quality circles (QC) according to regional ASHIP and ratings of current and future importance ASHIP Practising physicians QCs Participants in QC (%) Current importance 1 Future importance 2 Bavaria Berlin Brandenburg Bremen Hamburg 3 Hesse Koblenz Mecklenburg-Vorpommern (18.4) 60 (1.0) 450 (14.9) 306 (26.4) 328 (11.7) 1500 (18.8) 490 (28.0) 160 (6.8) - = Lower-Saxony 3 North Baden North Rhine North Wurttemberg (25.4) 700 (18.9) 3024 (22.4) 650 (11.8) = p Palatinate Rhinehesse 3 Saarland Saxony Saxony-Anhalt 3 Schleswig-Holstein South Baden South Wiirttemberg 3 Thuringia Trier Westphalia-Iippe (30.0) 187 (22.3) 310 (22.1) 250 (4.5) 82 (2.6) 955 (26.4) 600 (17.8) 152 (5.8) 370 (12.2) 375 (52.1) 1908 (17.5) - = = Total (17.0), Very low;, low;, moderate;, high;, very high.! =, Stable;, increasing;?, no prognosis given. 1 Estimates of participants because no exact figures available. Documentation and evaluation of QC by the regional ASHIP The majority (19 of 23) of ASHIP carried out systematic documentation to collect specific information about QC known to them, while the remaining four were planning to start documentation activities in the near future (Table 3). Comparative evaluation based on systematic documentation of individual QC had been established in 10 ASHIP, of which seven performed a more comprehensive evaluation (which included additional information besides subject, meeting intervals, number and specialities of regular and present members, minutes and results). This enabled comparisons of methods used, developments and results achieved by the different QC within a region. In nine ASHIP, these evaluations took place on a regular basis, in one only sporadically. Another five ASHIP were planning to start comparative evaluation. Seven of the 23 ASHIP demanded lists of participants from the moderators as a prerequisite for official recognition and grant support. Five other ASHIP delegated documentation and evaluation to an external scientific institute [11,12]. 38

5 Quality circles in ambulatory care Table 3 Systematic documentation of QC work by regional ASHIP Type of documentation Use of standardized forms to document existence of QC (subject, moderator, number of participants, frequency of sessions, chosen methods of data collection) and standardized documentation of individual QC meetings (e.g. date and location, present members, and short description of course of meeting) Use of standardized forms to document existence of QC only Use of standardized forms to document individual QC meetings only Gathering information in a formless way Standardized documentation planned in future Total No of ASHIP Facilitation and support of quality circles The regional ASHIP pursued different strategies to support and facilitate the work in QC. Material and organizational support were generally supplied, but specific facilitation by means of supervision, methodological or scientific advice, or the promotion of communication between different groups was only rarely provided. Follow-up meetings for moderators took place in 14 ASHIP. In seven, meetings were organized on a regular basis and in seven, on demand. On average 41.6% (23 95%) of the moderators (in total, 2420 moderators were invited) attended follow-up meetings, which were held approximately once or twice a year. Another five ASHIP stated that they were in the process of preparing meetings for their QC moderators. Four ASHIP had not organized any follow-up meetings by the qualifying date. The ASHIP encouraged the work of QC by providing organizational (22) or financial (20) support, materials (20) or mediation by resource personnel (16). The common goal was to motivate individual care providers to participate. ASHIP were of varying opinions regarding the acceptance of grants from drug companies to support QC and QI. Supplementary telephone interviews showed that a number of ASHIP apparently believed that QI should be protected from potential conflicts of interest, while others were not concerned about this. Eleven ASHIP reported the receipt of grants from drug companies in support of ambulatory QI activities, while the others denied such support. To receive financial compensation from the ASHIP the moderators had to meet varying requirements. In four ASHIP they were asked to provide structural data about the QC on a single occasion, and for general information about the meetings regularly. In addition to these basic data, two ASHIP required detailed information about the QC (e.g. methods, specialities of participants), while two others demanded detailed information about the meetings (e.g. self audit of participants, results). Nine ASHIP required moderators to provide all of these data. Although there seemed to be no direct co-operation concerning QC work, eight ASHIP noted support of QC by professional associations. Future perspective The ASHIP were asked to rate current and future importance of QC for QI in ambulatory care: 'How do you judge the current importance of QC work as a quality assurance measure in ambulatory care?' (five-point Likert scale from: 'very low' to 'very high'); and 'How will the importance of quality circle work presumably evolve?' ('increase', 'remain stable' or 'decrease? 1 ). Columns 4 and 5 of Table 2 illustrate that the majority of ASHIP rated the current importance of QC as 'moderate' (11), 'high' (7) or 'very high' (2). Only a few ASHIP estimated their current importance as 'low' (2) or 'very low' (1). Although these ratings were high overall, the future importance of QC was predicted to 'increase' (18) or 'remain stable' (4). None of the ASHIP predicted that the significance of QC would decrease (missing =1). Discussion Our survey permits a comprehensive overview over the development of QC in German ambulatory care based on data provided by QI co-ordinators in the regional ASHIP. Comparable data for some ASHIP available from our own system of continuous documentation [11] showed very good agreement. It is, however, reasonable to assume that validity is related to the comprehensiveness of the documentation and evaluation by the ASHIP. In Germany, quality assurance - commonly understood in the narrow sense as providers guaranteeing good quality health care is expected of physicians by health insurance funds and required by legislation. Nevertheless, in ambulatory care, its main structures are based on the voluntary initiative 39

6 F. M. Gerlach et al. of the physicians concerned. Our data show that this approach to QI in Germany developed successfully between 1993 and Within this period, the number of QC increased more than 100-fold, with approximately one sixth of ambulatory care physicians currently involved. The majority of ASHIP believed that the number and importance of QC would continue to increase. Although the QI co-ordinators who responded might be biased to view this development rather favourably, the fact that this opinion is supported by recent surveys carried out in two ASHIP regions suggests that 'socially desirable answering' is not an explanation for these results. According to these surveys, more than one-half of all ambulatory care physicians were, in principle, willing to participate in QC, about one-third was uncertain, and only every tenth physician completely rejected participation in peer review [13]. The increase in quantity of QC is, prima facie, a success, particularly considering that in Germany involvement in peer review is a relatively new element of QI in ambulatory care. Physicians are interested in better understanding and thereby improving their own practice. Particularly in Germany, working in QC gives ambulatory care physicians an opportunity to overcome the isolation due to predominantly solo practice and to improve professional co-operation. Participation is explicitly voluntary and no formal pressure or sanctioning is exerted on physicians to take part in QC. As the number of training courses for moderators has continued to increase, with more than two-thirds (68%) of the courses having taken place during the 12 months preceding our survey, a lasting development can be expected overall. Activities for support and facilitation differ markedly between the regional ASHIP. Major problems in most of the regions seemed to concern supervision of moderators, provision of methodological competence and structured exchange between different groups. While the pharmaceutical industry commonly supports continuous medical education in Germany, a number of ASHIP contain its influence in the field of QI and QC. The relative number of QC differed between regions and was lower in East Germany. Comments and additional interviews with QI co-ordinators in ASHIP made it clear that the extent of individual enthusiasm and support explained many of the differences. Physicians from the former German Democratic Republic found the term 'circle' problematic, as it reminded them of certain Marxist Leninist groups in which they had been forced to participate during the former regime. Another likely reason for the lower participation rates in East Germany was the profound transformation of the ambulatory care system after German reunification, which involved the dissolution of former outpatient community centres ('Polikliniken'). Most of the physicians in ambulatory care therefore had to found solo practices. This occurred at a time when the economic conditions of physicians in all of Germany were becoming increasingly difficult. Cost containment and the often difficult economic situation of the practices as well as additional legal regulations were perceived by many physicians not as a stimulus, but as an impediment to QI activities. Although the federal ASHIP has issued guidelines defining criteria that should be met by QC [3], our analysis of supplementary questions made it clear that the type of group recognized as a QC varied amongst the regional ASHIP. For instance, in three ASHIP it was difficult to make a definite distinction between Balint groups or continuing medical education (CME) groups (e.g. pain therapy) on the one hand, and QC on the other. On average, more than two-thirds of the existing groups were considered to meet the requirements for QC by the ASHIP. Most of the remaining groups seemed to be traditional CME groups not evaluating actual care. This raises the question of whether better evaluation, support, and facilitation could improve the content and thereby the impact of QC work. Quantitative data, such as that collected in this survey, reveal little about the detailed structure, objectives, and methods used by the existing QC. While documentation is performed accurately in most of the regions and often made a condition for compensation to the moderators, evaluation of QC activities is still underdeveloped. According to our survey, a minority of 10 ASHIP were performing some form of comparative evaluation of QC. Only five ASHIP were planning to start evaluation in the future. German physicians may perceive documentation and evaluation of QC activities as a threat to their autonomy when carried out by the ASHIP. Obligatory documentation of QC work as a basis for evaluation, moreover, often raises even the moderators' concern about external control, and more importantly, fear of additional paperwork without tangible results. Since, however documentation and evaluation are basic principles of systematic and continuous improvement [2], the overall lack of ongoing evaluation currendy presents a major deficit in many regions. Conclusions and recommendations Our survey shows that the establishment of ambulatory QI through voluntary QC has progressed rapidly since Given the evidence that QC are an effective method to improve patient care [14 17], this is a most encouraging development, which can be seen as a first step in the implementation of principles of continuous QI in German ambulatory care. It is important to emphasize that this step occurred mainly through voluntary initiatives of physicians with only subsidiary support of the professional bodies. However, further problem-oriented research will be needed to assess the more sophisticated aspects of effectiveness and the impact of individual QC on the quality of care specifically in Germany. Improved documentation and evaluation of this QI approach is necessary to achieve more accurate assessment of the development of different groups and of the methods used by them, and to evaluate the impact on professional development and medical practice. Because it is important to be able to describe the various stages of development of QC as well as to make comparisons between different groups and regions, our research group recently developed a 40

7 Quality circles in ambulatory care standardized instrument to document QC and their work. As an independent, external institution, we established a continuous documentation system with regular feedback reports for four ASHIP [11]. This could serve as a model for more widespread evaluation. In addition, we developed a more detailed instrument for QC assessment [18] according to the method of the Oxfordshire Medical Audit Advisory Group [19]. Our present results suggest that in order to achieve improvement of QC content and methodological competence increased support and facilitation is necessary from ASHIP, professional associations and independent scientific institutions. Because moderators often feel left alone to cope with difficulties in founding and guiding peer review groups, follow-up meetings in particular could be used increasingly to exchange experiences and transmit knowledge and skills. To support contacts between interested colleagues in Germany, Austria and Switzerland, we developed a database called 'Infoservice Quality Circle', in which moderators can register voluntarily and obtain information about other QC working either in the same area of interest or with similar methods. Further diffusion of experiences to and instruction of QC moderators and members are needed to promote problemoriented evaluation of the (potential) impact of the individual QC activities on health care [2]. We suggest that, in the future, support of QI activities such as QC should be based on a multifaceted assessment of the level of development attained by individual groups, and that this assessment should be performed by different parties. The ASHIP, being the contractual partner of the sickness funds in billing issues as well as having an obligation to promote QI, should be responsible for funding and for providing a legal framework and material support, and independent institutions should provide training and supervision, comprehensive evaluation and ongoing facilitation. Application of the principles, problems and solutions discussed in this article to specific regional or national situations might be helpful for similar QI activities in other countries. Acknowledgements We gratefully acknowledge the active participation of the staff of the German Associations of Statutory Health Insurance Physicians who made this survey possible. Comments on earlier drafts were provided by Wiebke Hellenbrand. Part of this research, with earlier results, was presented at the first National Forum of Physicians on Quality, January 1997, Berlin. References 1. Federal law on assurance and structural improvement of statutory health insurance (in German). Bundesgeset^blatt 1992; 59: Grol R., Lawrence M. Quality Improvement by Peer Review. Oxford: Oxford University Press, National Association of Statutory Health Insurance Physicians (Kassenarztliche Bundesvereinigung). Guidelines of the National Association of Statutory Health Insurance Physicians on procedures for quality assurance according to 135 par. 3 of the Social Security Code vol. 5 (in German). Dt. Artgebl. 1993; 90: B Bahrs O., Gerlach F. M., Szecsenyi J. (eds). Quality Circles in Health Care. Handbook for Physicians in Ambulatory Care (in German). Koln: Deutscher Arzte-Verlag, Bahrs O., Gerlach F. M., Szecsenyi J. (eds). Quality circles in Health Care. Handbook for Physicians in Ambulatory Care (in German). 2nd edn. Koln: Deutscher Arzte-Verlag, Herzog U. Activities in quality circle work (in German). Niedersdchsisches Ar^teblatt 1995; 68: Gerlach F. M., Bahrs O. Quality Assurance by Quality Circles in General Practice. Strategies for Implementation (in German). Berlin Wiesbaden: Ullstein Mosby, Szecsenyi J., Gerlach F. M. (eds). State and Future of Quality Assurance in General Practice. National and International Perspectives (in German and English). Stuttgart: Hippokrates, Szecsenyi J., Andres E., Bahrs O. et al. Evaluation of a training program for moderators of quality circles (in German). Z. Ar^tl. Fortbild. 1995; 89: Harter M., Vauth R., Tausch B., Berger M. Goals, contents and evaluation of training courses for quality circle moderators (in German). Z. Ar$L Fortbild. 1996; 90: Gerlach F. M., Beyer M. A new concept for continuous documentation of quality circle development in ambulatory care: Initial results from an information system in Germany. Qual. Health Can 1998; 7 (in press). 12. Tausch B., Harter M. Quality circles in primary care. Evaluation of a pilot project of the regional Association of Statutory Health Insurance Fund Physicians South Baden (in German). Munchen: Arcis, Gerlach F. M., Beyer M. Quality assurance in practice. Results of an analysis of needs and wants of ambulatory care physicians in Bremen and Bremerhaven (in German). [Report: AQUA- Materials No. Ill] and: Gerlach F. M., Beyer M. Quality assurance in practice. Results of an analysis of needs and wants of ambulatory care physicians in Saxony-Anhalt (in German). [Report: AQUA-Materials No. IV]. AQUA, Gottingen, 1996 and Grol R., Mokkink H., Schellevis F. G. The effects of peer review in general practice./. R. Coll. Gen. Pract. 1988; 38: Grol R., Wensing M. Single and combined strategies for implementing changes in primary care: A literature review. Int. ]. Qual. Health Care 1994; 6:

8 F. M. Gerlach et al. 16. Hartmann P., Bott U., GruBer M. et al. Effects of peer-review groups on physicians' practice. Eur. J. Gen. Pract. 1995; 1: Russell I. T., Addington-Hall J. M., Avery P. J. et al. Medical audit in general practice. I: Effects on doctors' clinical behaviour for common childhood conditions. Br. Med. J. 1992; 304: Beyer M., Gerlach F. M. Assessment of quality circles in ambulatory care. In Bahrs O., Gerlach F. M., Szecsenyi J. (eds). Quality Circles in Health Care. Handbook for Physicians in Ambulatory Care (in German). 4th edn. Koln: Deutscher Arzte-Verlag, 1998 (in press). 19. Derry J., Lawrence M., Griew K. et al. Auditing audits: The method of Oxfordshire Medical Audit Advisory Group. Br. Med.]. 1991; 303: Received in revised form 30 June