Received: 19 July 2009; Accepted date: 21 November 2009
BackgroundThe Quality and Outcomes Framework (QOF) has had a major impact on chronic care provision in British general practice. Various countries are looking at whether a similar initiative could be used in their primary care systems. An extensive quality indicator system like the QOF does not exist in German general practice. AimTo describe and explore the views of German general practitioners (GPs) on the clinical indicators of the QOF. MethodsQualitative study based on focus group discussions and a framework approach for data analysis. Fifty-four German GPs were involved in seven focus groups in German primary care practices. Results German GPs expressed mixed views regarding the validity of the QOF clinical indicators to measure the quality of primary care. Most thought that these indicators covered areas that were relevant for German general practice and which were only partially covered by German quality initiatives. Participants had mixed opinions regarding linking pay and performance. Many thought that in deprived areas it would be difficult to achieve targets. Exception reporting would make achieving these targets easier, however, some believed it could lead to manipulation of figures. Many GPs saw QOF clinical indicators as a helpful structure, yet feared that introducing something similar would increase the administrative workload and be a threat to patient-centred care. Many participants were anxious that a QOF-like system could be influenced by sickness funds or the pharmaceutical industry. A few feared data protection problems if such a system were to be implemented. Several GPs expressed concerns on who would set and control such quality initiatives, feared for their autonomy and expected that in the future similar systems would be imposed upon them.ConclusionsParticipating German GPs had various concerns regarding the QOF clinical indicators and the idea of implementing a system like the QOF in German primary care. These concerns were mainly related to the validity of the indicators, the link between pay and performance, structured care versus patient centredness and the fear of external influences.
family medicine, focus groups, qualitative research, quality indicators
Improving the quality of medical care has become a major issue for all healthcare systems.1 However, there are many barriers to changing clinicians’ behaviour and most initiatives to improve patient care have had only limited or mixed effects.2,3 From that point of view the QOF seems an exception. The QOF is a pay for performance system, consisting of clinical, organisational and other indicators. The clinical indicators represent the main part of the framework and cover problem areas ranging from coronary heart disease and depression to obesity and smoking. The QOF was introduced in theUK in April 2004 and although it is a voluntary scheme, 99.8% of UK practices participate. Average achievement has consistently been over 90%, with a mean score of 96.8% in 2007 to 2008.4,5 What does this mean in relation to quality improvement in UK primary care? Several early studies on the QOF compared data from just before and just after its introduction and suggested that the framework had improved the quality of care for conditions like asthma, diabetes and heart disease.6–10 Other early research indicated that the differences in QOF achievement between affluent and deprived areas were small and that the QOF may have diminished health inequalities. 9,11–15 However, more recent studies suggested more negative outcomes. The QOF seemed to have improved the quality of chronic care, but the pace of improvement was not sustained once quality goals were reached. Further, the QOF seemed to have had a negative effect on the quality of care of conditions which were not included in the QOF and may have reduced continuity of care.16 Finally, another recent study indicated that the scheme failed to capture a significant proportion of diabetic patients and as such may not have been as efficient in reducing inequalities in diabetes care as initially was hoped.17
Despite these recent concerns the QOF is seen as a major innovation amongst quality improvement measures and various countries are looking at whether a similar initiative could be used in their primary care systems.18,19 At an international level there is some evidence that the development of quality indicators follows a similar direction and that a transfer of such indicators between countries is possible, albeit with caution.20–22
In German general practice there are quality initiatives such as clinical guidelines and a few disease management programmes (DMPs) in which individual patients can opt to participate.23 However, at practice level, an extensive quality indicator system like the QOF does not exist and as such there seems to be a ‘quality indicator gap’. Before such a system could be developed or transferred to German primary care it would be necessary to know the views of the clinicians who would be expected to use it. Therefore the aim of our study was to describe and explore the views of German GPs on the clinical indicators of the QOF.
Preparation and planning
We based our methodology on a phenomenological approach. There were two main reasons for this choice. Firstly, we wanted to describe and explore the subjective views of German GPs; secondly, we recognised our restrictions, especially in relation to manpower and resources.
In consequence we wanted to use focus groups to collect data with the aim of capturing opinions and to using interaction to spark new ideas. We decided to use the full set of QOF clinical indicators (2006 version) as a guide for our participants, as this set forms the main part of the QOF and represents the key principles. We thought that the whole QOF would be too long and less relevant to German GPs and therefore not feasible to discuss in a focus group. The aforementioned guide was translated into German. We also developed a semi-structured topic guide.
First we pilot tested our focus group procedure by organising two sessions involving a total of seven GPs known to the research department. Based on these we slightly adapted the topic guide, while the QOF clinical indicators guide remained the same.
Our sampling strategy was to recruit participants via German GP ‘quality circles’ (peer review groups). These peer review groups consist of GPs who meet up approximately once every two months. At each of their meetings they discuss clinical topics or practice issues as a means of quality improvement and continuing professional development (CPD). As we wanted to collect the views of a broad spectrum of German GPs, we phoned the coordinating GPs of urban and rural peer review groups in a wide geographic area in the northwest of Germany. Those coordinating GPs who were interested in the project were sent an email or a fax with extensive information regarding the study. In total we contacted 15 of these coordinating doctors and seven of their peer review groups agreed to participate. Seven focus groups with 54 participants (34 male and 20 female German GPs), covering practices in the north-western part of Germany, were conducted during the period from June 2008 to November 2008. Focus groups included six to 11 participants and involved GPs from urban and rural practices.
Data collection and analysis
Coordinators of the peer review groups determined where the focus groups took place, usually in a primary care practice. At the start of each session every participant received the QOF clinical indicators guide and everyone was given ample time to read and study this before the focus group discussion. One researcher with extensive experience as a GP in British primary care (HVDH) moderated the sessions using the pilot tested semi-structured topic guide. Duration of the focus groups was between one and a half and two hours. All discussions were audiotaped and transcribed verbatim.
In line with the methodology we applied a framework approach for analysis.24 Two researchers (HVDH and SH) independently familiarised themselves with the data and coded the transcribed interviews (using Maxqda 2007 data analysis software, inductive mode) to identify a thematic framework.25 They matched their provisional frameworks, this matched version was shared with the other two researchers (PM and EHP), and all were involved in finalising the framework. The coordinating GPs of the participating peer review groups were invited to provide feedback on this analysis and no alterations were required. Finally, all four researchers took part in drafting the paper which, via iterative discussion over many weeks, led to a consensus article. A qualified medical translator translated quotes from German into English.
Our focus groups revealed four main areas of concern related to the idea of implementing a system like the QOF in German primary care: validity of quality assessment in primary care, implications of linking pay and performance, indicator orientated care versus patient centredness and fear of external influences.
Validity of quality assessment in primary care
German GPs expressed mixed views on whether the QOF clinical indicators captured quality in primary care. Many saw the collection of QOF clinical indicators as a ‘set of clinical guidelines’. Although most GPswere positive about clinical guidelines, they thought these had limitations in covering quality, especially related to the care of patients with multiple problems and conditions.
‘It’s a big problem that there are many elderly patients, who have coronary heart disease as well as asthma, and then diabetes too, and perhaps they are also obese or depressed or whatever, and then I am sitting there and I have to get through four of such schemes (QOF), that is really too much, no way.’ (Focus Group (FG) D, GPA, 16)
The GPs also had mixed opinions on whether the ‘QOF percentages’ (the proportion of patients for whom a predefined target was met) could be used to measure quality. Some felt that their use was a good approach to making the essentials quantifiable. However, many participants thought that these percentages would shift the focus too much towards easy to measure parameters.
‘And the risk is, in line with public health interests, that more and more those diseases are measured, or treated, or are in the spotlight, which can be measured with hard data! Since these can easily be processed into some programmes, you can simply draw straight lines, and you also can define demands or make adjustments on pay and so on ...’ (FG H, GP A, 76)
In general, GPs saw timeframes as good tools for measuring quality. It was noted that in the QOF intervals of 15 months were often used. Many considered this an odd time span, questioned the statistical and practical value and thought it could promote minimal standards. Some believed that the interval length should depend on the weight of the parameter.
Most GPs thought that the QOF clinical indicators covered clinical areas that were relevant for German primary care. Many thought that they did not cover German general practice in full. Several thought that a wide range of areas, from low back pain to unspecific problems like ‘dizziness’, were missing and some expressed the opinion that there was too much emphasis on managing disease as opposed to health promotion.
‘And what’s missing is the real big issue of minor illnesses, isn’t it? Prevention, pre-morbidity, health education, recognising not just the disease but also diagnosing health and thus patients’ education, behavioural change, these are actually the crucial issues in my practice.’ (FG I, GP C, 19)
Most participants mentioned that the QOF clinical indicators had some overlap with German DMPs but thought that the DMPs were more bureaucratic and covered a much smaller area of primary care. Finally, several GPs mentioned that they had first been sceptical about the implementation of the DMPs, but thought thatwith time they had become more efficient in using them, and now also saw some value in this system.
On being asked what would be an incentive to use a system like the QOF, many GPs mentioned payment. There was debate regarding how much extra money would be required for these participants to use such a system. However, several participants felt ashamed of accepting pay for performance and had doubts regarding the principle. They mentioned that they would also use a scheme like the QOF if they were convinced that it was beneficial for their patients, in line with professional guidance and restricted to a few relevant key issues (Box 1).
Many GPs had doubts whether practices could be compared based on QOF figures and thought that GPs in deprived areas would find it hard to achieve targets, despite putting in a lot of effort.
‘I’ll give Aspirin tomypatient with coronary heart disease, but I cannot influence it, for example, when HbA1c is used as a parameter for diabetic patients. If I’ve got a certain practice population, I can break my back, but I’ll never get anywhere. That’s the problem.’ (FG C, GP A, 88)
As such they considered the link between pay and performance to be unfair.
‘Exception reporting’ is the procedure in the QOF that allows certain patients to be excluded from the percentages (for example, patients who clearly express that they do not want indicated medication or tests). This procedure was discussed in the focus groups. Many participants felt that exception reporting would make the link between pay and performance more acceptable, but some feared that it could lead to manipulation of figures, for example by excluding the ‘non-compliant patients’.
Many GPs saw the QOF scheme as a helpful structure for systematic care and professional reflection, and as a potential quality marker. Most participants wanted to receive feedback on their treatment, as long as itwas adequate and timely, and for that reason wouldwelcome a system like the QOF. They considered this to be missing in their current system.
‘The positive point regarding these systems is that one gets a reflection on the job done, and also obtains figures for comparison with other practices.’ (FG G, GP B, 155)
‘If I get feedback that I have improved, that I do well, ... that I ampositively motivated, that I amconvinced of the purpose, that it is well delivered, well, then I think it would be very valuable.’ (FG I, GP C, 128)
Some thought that a lot of work related to the QOF clinical indicators could be delegated to the ‘practice assistants’, making their job more interesting while at the same time saving some of the doctors’ time. However, the majority feared that theQOFwould lead to an increase in administrative workload. Related to this, a lot of GPs expressed a dislike for working with percentages and statistics. Many recognised information technology (IT) as the key factor in addressing this issue yet felt that the current practice IT systems in Germany were limited and that major alterations would be required in order to be able to use a system like the QOF.
Many participants saw the indicator orientation as a threat to patient-centred care and worried that it could lead to a loss of patient orientation, to ‘tunnel vision’ and to a loss of use of GP intuition. This could go so far that non-compliant patients would be seen as ‘the enemy’ of their performance figures (Box 2).
Several GPs were concerned about who would set and control quality initiatives like the QOF and feared for their autonomy. Many participants did not like the idea that QOF data could be accessed and collected by a third party, such as a health authority. Some were concerned that quality indicators could be manipulated by these third parties, for example the pharmaceutical industry or the sickness funds. Related to this, a few GPs thought there would be data protection issues if a system like the QOF were implemented in Germany (Box 3).
Finally, various GPs expected that in the future similar systems would be imposed upon them.
‘When we get such systems, once more you get the feeling that there is again such a bureaucratic system, which demands issues and requires fulfilling these demands, what you don’t want to do ... However, I think it is always such a dilemma; on the one hand it is obviously good to have structured care, on the other hand: when this suppresses everything, then it won’t lead to anything.’ (FG G, GP I, 158)
Most German GPs thought that the QOF clinical indicators covered areas that were relevant for German general practice and that these were only partially covered by German quality initiatives. Most participants expressed ambiguous feelings regarding the QOF clinical indicators and the idea of implementing a system like the QOF.
A particular strength of the study is that for the first time German GPs have been asked to provide their views on the QOF clinical indicators. Another strong point is the relatively large number of GPs, rural and urban, that participated in the focus groups.
The study has several limitations. First the participants were recruited from peer review groups.As these groups consist of GPs interested in quality improvement and CPD, they may be more positive about quality indicators than GPs who are not participants, which could have caused bias. Also, they were GPs from the north-western part of Germany and views of colleagues in the south and east may be different. Finally, the views of the GPs were expressed based on their assessment of the indicators and this may not be the same as what they would do in practice.
At the introduction of the QOF in 2004 British GPs questioned whether the scheme would improve the quality of care and feared especially a reduction in their autonomy, less patient centredness, a reduction in the continuity of care and an increased administrative workload. Also it was expected that the framework would show a poor performance in deprived areas and that it might lead to manipulation of figures. However, subsequent research suggested that theQOF had improved the quality of chronic care, although this did not improve further once quality targets had been achieved. The scheme seemed to have had a negative impact on the quality of care of nonincentivised conditions and it may have reduced the continuity of care. After using the framework for several years UK GPs expressed the view that QOF had changed the way of working in their practices. For example, they thought that practice nurses had a more central role in chronic care and that care provisionwas more fragmented. Also it was noticed that at times there was a tension between the patient’s and the QOF’s agendas. The feared extra administrative workload, loss of autonomy, problems related to poor performance in deprived areas and the risk of manipulation of figures did not seem to have materialised. 6–17,26–29
Many concerns from British GPs before the introduction of the QOF were similar to those identified in this study. In contrast, this research project revealed some differences from the findings of the literature around QOF. For example many German GPs feared the influence of third parties, like the pharmaceutical industry or sickness funds, and some mentioned data protection problems as an issue.
This study has provided some insight related to what German GPs think regarding QOF clinical indicators. As these indicators were thought to be relevant and as there seemed to be a ‘German quality indicator gap’, one could consider whether certain parts of the QOF clinical indicators could be used in German primary care. However, the focus groups revealed that there were several areas of concern related to the idea of implementing a system like the QOF.
Some of these concerns have not been mentioned before in the literature in relation to the QOF and perhaps are specific to Germany, for example concerns related to fear of the influence of the pharmaceutical industry. Related to this the specific structure of the German healthcare system needs to be taken into account with different players, such as GPs, sick funds and the industry representing different interests. Other concerns were similar to previous British ones, for example those related to the expected underperformance in deprived areas. Several of these British worries have not materialised in the UK, but that does not necessarily have to be the case in Germany.17,18 Therefore if one were to consider implementing a system like the QOF in Germany, it would seem important to take all of the identified concerns seriously and to take all of them into account. Perhaps most worrying was that several German GPs expressed concerns regarding who would set and control the quality initiatives and worried that these would be imposed upon them. It appears important to address this matter, for example by organising meetings involving both policy makers and clinicians, to try and overcome differences around these initiatives. After that, if one still wanted to follow a similar direction as the QOF, one could think about piloting similar schemes in a few GP practices before implementing them on a wider scale, or one could even consider developing a ‘local QOF’ to address local concerns.4
What would be the views of German patients regarding a system like the QOF? This question could not be answered in this study. Limited information exists regarding the views of British patients on the QOF. A recent study showed that British patients and their GPs had different views on the depression questionnaires used in QOF.30 It seems important to get the views of German patients on such systems if these are considered for development or implementation, for example by inviting German patient groups to bring forward their ideas on priority subjects.
Overall the debate on quality indicators forGerman primary care has only just started and more research is required regarding the development and implementation of quality indicator systems in German general practice.31,32 This paper may make a small contribution to this debate.
We thank all participants for their cooperation. Special thanks go to T Jones for his support, to S Friedrich and M Ferna´ndez-Belver for their help related to the translations and toD Lawrence for secretarial assistance.
This work was done at the Institute of General Practice of Hanover Medical School, which is state funded.
Received from the Ethics Committee of Hanover Medical School.
Not commissioned; externally peer reviewed.
All Published work is licensed under a Creative Commons Attribution 4.0 International License
Copyright © 2019 All rights reserved. iMedPub LTD Last revised : May 24, 2019