Deputy Director Postgraduate Research Education, School of Nursing Postgraduate Division, University of Nottingham, UK
Richard Baker OBE MD FRCGP
Head, Department of Health Sciences, University of Leicester Medical School, UK
Received date: 15 November 2006; Accepted date: 24 February 2007
This paper discusses the concept of measuring organisational culture in primary care and reports on the construction and early pilot work surrounding the development of the Practice Culture Questionnaire (PCQ). The PCQ was designed specifically to identify variations in resistance culture to quality improvement activities in UK primary care teams. Questions were derived from a model of clinical governance and from interviews of primary care teams about quality improvement activities. A 25-item measure was devised and piloted with a sample of 21 practices. Response rates were satisfactory and the findings discriminated between primary care teams. Further evaluation of the measure is now required. The value of a measure that identifies resistant culture to quality improvement in primary care teams is discussed.
Over the last 20 years, organisational culture has generated significant research interest within a variety of academic disciplines including sociology, manage-ment science and psychology. The way these discip-lines approach the construct of organisational culture is often radically different, as there is little agreement amongst them on precisely what the concept is. The division in agreement about what organisational cul-ture might be is mirrored in the debate about whether it is possible to measure it and if so how. Qualitative researchers argue against a quantitative research methodology on the grounds that it misses out the rich nature of culture.[2,3]
Quantitative researchers on the other hand argue that in order to investigate culture in different organ-isations, standard measures must be developed.[4,5] This latter approach reflects a management science perspective, implying that culture is an attribute that organisations have and that it can be identified and changed.[6,7] This idea has been challenged by sociol-ogists and anthropologists who argue that the culture of an organisation does not exist in the sense that it can be identified and changed, but exists only in the sense of it being a conceptual value system. The differences are only in the degree to which the values are socially constructed and shared by people in an organisation. Organisational culture as a measurable concept in UK health care was promoted by Liam Donaldson in a series of publications that stressed the importance of overcoming a ‘resistant to quality improvement cul-ture’ in UK healthcare organisations.[8–10] As the new Chief Medical Officer and architect of clinical govern-ance, he urged healthcare managers to recognise that quality improvement was an organisational issue. Organisations in his view needed to embrace the new modern NHS and create a culture of continuous quality improvement.
When discussing organisational culture in primary care he presented a list of qualities that he believed were associated with a culture that was resistant towards a culture of quality improvement.
His list was drawn from a review of the business and healthcare literature and is reproduced in Box 1.
Here then were 12 qualities of resistance contributing to a culture that the Chief Medical Officer in 2000 felt were shared by healthcare organisations – including general practices – that would hold back develop-ments in provision of high-quality care. The qualities that Donaldson identified were not isolated aspects of individual behaviour but very much linked to team-work and organisational culture and supported by other writers at the time. Donaldson was keen to emphasise in his writings that clinical governance was essentially an organisational concept. He acknowledged that the introduction of clinical governance and the creation of primary care trusts would push organ-isational culture to the top of the healthcare agenda.
The importance of primary care organisations working cohesively within a framework with clear goals and objectives directed towards quality improvement will be one of the keys to the success of the initiative.
Donaldson’s view of organisational culture, by his own admission, was drawn from the business and marketing literature. An organisation’s culture has been defined in the business and marketing literature as the ‘complex set of values, beliefs, and assumptions that define the ways in which a firm goes about its business’. This view suggests organisational culture can be understood from accessing the shared beliefs about the workplace that exist between colleagues in the organisation. This includes the beliefs, attitudes, values and norms of behaviour within the organisation and how staff perceive these. Another similar ap-proach is to consider the way in which people share a common understanding of ‘how things get done’ in their organisation.[1,11]
Is there a measure of organisations in health care working cohesively together within a framework with clear goals and objectives directed towards quality improvement? Would such a measure be useful in helping identify organisations that have got such a framework and those that have not?
Edgar Schein suggests that culture in an organis-ation is a crucially important factor to a business, and operates at three levels: on the first level are the artefacts, the visible things which people do; second the espoused values that drive the strategies that impinge on the systems the organisation employs; and on the third level are the basic underlying as-sumptions which are the deeper unconscious beliefs held by employees about the organisation and how it should behave. Schein believes that a measure of an organisation’s culture can be found using well-thought out and appropriately worded questionnaires.
Although there are examples of questionnaires to measure culture in secondary care there are very few if any that apply to the working environment of UK primary care. A systematic review reported by Stevenson in 2003, using the search terms ‘measuring organ-isational culture’ and ‘questionnaire’ searching from 1966 to 2003 identified only eight published studies in health care involving the use of a specific organisa-tional culture questionnaire. The studies are listed in Table 1. A review of the studies reveals that only two of the studies involved UK primary healthcare staff.
The two attempts at measuring culture in UK primary care using a questionnaire are different in scope and size. The first study used a questionnaire developed from the established ‘learning organisation’ model. This study, however, concentrated only on staff in 15 practices and only on the employed staff (e.g. nurses and receptionists).
The study by Stevenson (2000) used a specially devised questionnaire, The Practice Culture Ques-tionnaire (PCQ) that focused on the views of the whole primary care team. The primary care team in this context is viewed as a ‘whole organisation’, including all those involved in the delivery of care. The questionnaire has, since that publication, been piloted on over 150 practice teams and over 1500 primary care staff.
This paper will explain how the PCQ was developed and then shown through pilot studies to be a reliable and valid measure of what Donaldson regarded as organisational culture in primary care.
The construction of the PCQ (see Appendix) was underpinned by the essence of the working definition of culture outlined above which was: ‘the shared set of values, beliefs and assumptions of the staff who work in the organisation ...’, and was developed with the following guidelines:
• the questionnaire should be easy to complete, e.g. using a simple easy to understand response format
• the questions should be based on how the respon-dent feels the practice team views various clinical governance activities
• the questions should be relevant to all grades of practice staff and relate to the sorts of clinical governance activities the practice could be expected to be carrying out
• the questions should reflect the way primary care staff think about quality improvement activities drawn from earlier reports of what staff had said during interviews
• the questionnaire should look easy to complete for time-conscious healthcare professionals (i.e. fit onto one side of A4 paper).
Since the questions were be based around how the respondent felt the practice team viewed various clini-cal governance activities, the questions were linked to the 12 critical aspects of a resistant organisational culture outlined by Donaldson. His 12 critical aspects, however, were too generic for the specific context of primary health care, and therefore the questions needed to be contextualised to primary care quality improve-ment activities so that they would be immediately recognisable to all UK practice staff.
For example, UK primary care staff would be more likely to be able to judge whether their practice carried out clinical audit than answering a general question about whether their practice was in favour of ‘moni-toring clinical performance’. Table 2 illustrates how each Donaldson attribute was ‘contextualised’ into a clinical governance quality improvement activity.
Each quality attribute was framed as a statement with an associated response in Likert-scaled format. The construction process ensured that the general quality concepts were covered using statements that focused on activities that were, or should have been, familiar to most practice staff.
Getting the wording right
Earlier interview studies with staff in general practice teams had already illustrated the kinds of statements that practice representatives had made quite freely about their practices. For example:
‘We [the practice] have considerable experience in audit.’
‘We regard audit as very much part of our activities.’
‘We’re supposed to be an innovative practice.’
Similar types of statement were forthcoming from primary care staff interviewed in an NHS Trent clinical governance baseline study:
‘We don’t feel we get much value from patient question-naires.’
‘Our reception staff aren’t that interested in going on courses.’
‘This practice takes patient complaints very seriously.’
These statements provided the springboard for con-structing PCQ statements in a similar fashion that staff working in practices would recognise and feel com-fortable agreeing or disagreeing with.
To increase the reliability of the scale, each of Donaldson’s 12 attributes was used as the basis for two separate statements. The 24 paired statements were then constructed. These became short unam-biguous statements about the practice, such that agreement or disagreement by the respondent would provide support for, or against, the view that the practice was positively, or negatively, disposed to the activities of clinical governance (Table 3). As the questions were being constructed, the notion of a blame-free culture emerged in NHS thinking about health care. Therefore, a 25th statement that asked about how the practice responded to staff mistakes completed the questionnaire items.
These questions that were created now linked dir-ectly to Donaldson’s 12 critical issues and also mapped onto the essential components of primary care clinical governance suggested by Roland and Baker, namely:
• being responsible and accountable for quality of clinical care
• involving all staff in quality improvement activity
• developing appropriate risk management pro-cedures
• developing appropriate procedures to identify and remedy poor performance.
This suggests a questionnaire measuring resistance to clinical governance built from Donaldson’s 12 critical issues would satisfy the initial requirements for con-tent and face validity. Further support for validity of the questionnaire items came from: discussions with GPs and practice nurses who helped fashion the questions to ensure they were appropriately worded for the intended audience; the acceptability of the questionnaire to pilot staff; and the relatively few questionnaires returned uncompleted from pilot studies.
Questionnaires about health practices are sometimes prone to social desirability where the respondent gives answers that he or she believes would be regarded positively. The PCQ items were specifically designed to minimise social desirability effects by focusing on statements about the practice the staff actually worked in and how, in Handy’s phrase, the respondent’s practice ‘went about its business’.
A methodology for distributing the questionnaires was devised. Once agreement to take part had been given, copies of the PCQ were provided to a practice for completion by the primary care team. The ques-tionnaires were sent to, and distributed to each mem-ber by, the practice manager and when returned to him/her (in a sealed envelope if required), forwarded to the author for scoring. The ‘practice team’ is therefore what the practice defined it to be. There is an argument that different practice organisations will view the membership of their primary care team differently. The pilot studies described below, however, tended to support the view that the distribution of grades of staff included by practices was largely similar individual’s view of their practice to be 100 points and an individual’s lowest total PCQ score for their view of their practice to be zero..
With 25 statements on a four-point scale, a scoring strategy was developed that scored each item from 0 to 4. A zero score on an item indicated the respondent’s view that their practice had a very poor attitude towards that aspect of clinical governance. For example, if they clearly disagreed that the practice was well organised or that they very much agreed that the practice regarded clinical audit as more trouble than it was worth. A score of 1 indicated the respondent’s view that the practice was not generally positively disposed to the item. A score of 2 indicated that the respondent was unsure of the practice position on the issue and a score of 3 or 4 indicated the respondent felt the practice was generally positively or definitely pos-itively disposed to the issue.
The fact that there were 25 items each with a possible maximum score of 4 and a minimum score of 0 provides the PCQ total maximum score for an Feasibility and acceptability pilot study
The first pilot study involved circulating 21 practices and was designed to
• check the acceptability of the PCQ by staff asked to complete it
• assess whether the practices could be differentiated by the average (median) score of the practice on the PCQ scale.
This first pilot study did support the feasibility of the PCQ as a useful measure of practice culture. The mechanism of distributing and collecting the ques-tionnaire through practice managers was successful, with an overall questionnaire return rate of 88%. In summary, 17 out of 21 pilot practices agreed to take the questionnaire; 5 out of 17 practices had 100% return rates against questionnaires issued. Only 3 out of 17 practices had return rates less than 50% against questionnaires issued (Table 4).
More importantly, the analysis of the pilot data by practice team indicated that the PCQ was also able to differentiate between practices with different attitudes towards clinical governance. The 17 practices that returned pilot questionnaires provided a practice aver-age PCQ score that ranged from 49 to 88. The range of scores within a practice was also noted as a useful indicator of how concentrated feelings were within a practice (Table 5).
Table 6 outlines how four types of practice might be identified through their team median score and their spread score.
Practices with high median score and low spread indicate a team that is comfortable with the principles of clinical governance and quality improvement ac-tivities. The low spread of scores within the team emphasises the cohesive nature of the team, and suggests they share information and working prac-tices.
Practices with low average PCQ score and low spread suggests a team who feel their practice is predominantly resistant to activities such as those suggested in the questionnaire. The low-spread score indicates that there is general agreement by all the team about the practice and its negative view of clin-ical governance and associated quality improvement activities. Questionnaire feedback could be used to establish which activities respondents feel that the practice is most resistant to and, if attitude change were sought, appropriate educational support strat-egies that could be considered.
A practice with an average culture score but high spread suggests that team members may hold mark-edly different opinions and beliefs about the value of clinical governance in their practice and the quality improvement activities associated with it. The way in which staff in these practices might be encouraged to change may differ, depending on who is most resistant to which particular aspects of quality improvement and to what degree.
This paper discusses a proposed relationship between organisational culture and resistance towards quality improvement activities in primary healthcare teams. The difficulties of measuring organisational culture in primary care are considered, and a description of the construction and initial testing of the Practice Culture Questionnaire has been given. The PCQ was designed specifically to identify variations in resistance culture to quality improvement activities in UK primary care teams. The early pilot work suggests that this approach and the PCQ itself could be useful to practices who are interested in measuring how well they are doing in encouraging their staff, as Donaldson suggests, ‘to work cohesively within a framework with clear goals and objectives directed towards quality improve-ment’.