Family Practice Vol. 16, No. 3, 305-311
© Oxford University Press 1999
Action research: a useful method of promoting change in primary care?
Division of General Practice, School of Community Health Sciences, The Medical School, Queen's Medical Centre, Nottingham NG7 2UH and
a Division of Child Health, School of Human Development, University of Nottingham, The Medical School, Queen's Medical Centre, Nottingham NG7 2UH, UK.
Hampshire A, Blair M, Crown N, Avery A and Williams I. Action research: a useful method of promoting change in primary care? Family Practice 1999; 16: 305–311.
Received 25 August 1998; Revised 8 January 1998; Accepted 28 January 1999.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Background. Action research has been used successfully to promote change in disciplines other than medicine, but there are few examples of its use in primary care.
Objective. We aimed to discuss the benefits and difficulties of using action research in primary care using the example of child health surveillance provision in general practice.
Methods. Twenty-eight general practices were randomly allocated into two groups. Action research was used to promote change in 14 practices by facilitating practice meetings and by providing written feedback. The other 14 practices received written feedback alone. The two groups of practices were compared using the following: (i) semi-structured interviews with one health visitor and GP from each practice; (ii) observation of baby clinics; (iii) questionnaires to parents; and (iv) return rates of child health surveillance reviews from the personal child health record.
Results. All 14 practices in the action research arm of the study met as individual practice teams and decided to make changes to their provision of child health surveillance. Ten practices audited their child health surveillance as a result. More health visitors in the action research practices than in the comparison practices reported changes to child health surveillance, audit, communication and use of the personal child health record. The majority of health visitors and GPs thought involvement in the action research process was beneficial. However, we were unable to show a statistically significant difference between the two groups of practices in baby clinic provision, parent satisfaction or the return rate of child health surveillance reviews.
Conclusion. Our study suggests that action research is a successful method of promoting change in primary care. However, measuring the impact of change is difficult.
Keywords. Action research, primary health care, quality of care.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The quality of care provided in health services is no longer taken for granted.1 –4 In order to improve health care, methods of changing doctors' behaviour have been explored, and so far there appear to be no easy solutions.5 –9
In education, industry and, more recently, in nursing, action research has been used to facilitate change and improve the services provided.10 –17 It is also considered to be a good method of bridging the gap between research and practice. However, there are few published examples of action research being used to improve primary health care.18,19
There has been much debate over the use of the term ‘action research’. It covers a spectrum of research activity and its definition varies in different contexts and in different countries. Most definitions of action research characterize it as: (i) focusing on change and improvement; (ii) involving practitioners in the research process; (iii) being educational for those involved; (iv) looking at questions that arise from practice; (v) being a cyclical process of collecting, feeding back and reflecting on data; and (vi) being a process which generates knowledge. For the purpose of this study, these aspects will be used to define action research. Other researchers have seen it as a means of empowering practitioners and changing the social context in which the research has been done.10 –17
In England and Wales, the 1990 contract20 encouraged a shift in the provision of child health surveillance (CHS) from child health clinics run by community paediatricians and health visitors to general practice. There was concern that GPs would not have sufficient training, skills or motivation to provide an appropriate service.21 –23 We evaluated child health surveillance provided by primary health care teams in general practice in one health authority in the UK. We used an action research model with the aim of facilitating change, improving service provision and developing a local standard for child heath surveillance.
The aim of this paper is to discuss the benefits and difficulties of using action research in primary care using our research as an example. In order to do this, we have outlined the range of methods used in our study and have provided an overview of the results.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Forty of the 120 general practices in the Nottingham Health Authority area were invited to take part in the study. Of these, 31 agreed to participate. One practice was excluded because a third of their patients lived in an adjacent health authority, another because the GPs did not provide child health surveillance, and a third because the practice disbanded during the first six months of the study.
Data on the CHS activities of the 28 study practices were collected by direct observation of practice baby clinics, questionnaires to parents, semi-structured interviews with one health visitor and one GP from each practice, and collection of carbonated copies of the CHS reviews recorded in the personal child health records (PCHRs) of all children born and subsequently registered with the practices.
In the UK, health visitors are usually employed by community trusts but work closely with GPs to provide care for pre-school children. The PCHR is used by most health authorities and is usually given to parents shortly after the birth of their child. It contains advice for parents and space for recording growth measurements, immunizations, CHS reviews and contacts with all health care professionals.
After 9 months of data collection, the practices were randomly allocated to two groups, an ‘action research’ group and a control group, stratifying for single-GP versus larger practices and the socio-demographic characteristics of the practice area.24
All the practices were provided with quarterly reports on their child health surveillance based on data collected from the PCHRs. These reports were anonymized but enabled individual practices to compare their ranked performance with other practices for return rate, legibility and problems detected at CHS reviews.
The action research practices were also visited at 4-monthly intervals by two researchers who facilitated practice team meetings. At these meetings, the practice teams discussed the feedback they had received and identified problems in their provision of child health surveillance and the data collection techniques used. They discussed how child health surveillance in their practice could be improved and decided what changes should be made. They also discussed the development of a local standard for child health surveillance. The meetings were minuted by one of the researchers and copies of the minutes, including the decisions that had been made, were sent to the practices shortly afterwards.
During the feedback period, data collection from the PCHRs continued. After the feedback year, at all practices, the baby clinics were again observed, questionnaires were sent to parents, and a health visitor and GP were interviewed. Where possible, the same health visitor and GP were interviewed before and after feedback. Finally, the results of the research project were discussed with members of most of the practice teams at an away day/ workshop.
The minutes of the practice meetings and the open response questions used in the interviews and parent questionnaires were coded and categorized using the constant comparative method.25,26 The quotes used represent the most frequently expressed views of responders. Quantitative data were analysed using SPSS 6.1 and Epi info 6. Chi-square tests for significance are quoted with Yate's correction and Fisher's exact probability test, where appropriate. Correction for the effect of clustering has been made in the analysis of the questionnaires to parents because randomization occurred at practice level. An intra-cluster correlation coefficient of 0.05 was used.27,28 Stata version 5 was used for Poisson regression analysis of the return rates per practice of the child health surveillance reviews.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
The study practices were representative of all practices in Nottingham Health Authority for Jarman deprivation scores29 and for proportion of practices that were fundholding, single-hand or who owned their practice premises (Table 1
). The study practices were significantly more likely to be registered for child health surveillance (chi-square = 3.2, P = 0.04) and to be involved in teaching (chi-square = 8.1, P = 0.004).
|
The facilitated practice meetings
Five of the action research practices had regular practice meetings, and at 4-monthly intervals we facilitated these, to discuss child health surveillance. Organizing meetings at the other nine practices was a lengthy process, with the practice manager or health visitor being key contacts. All meetings were held in the practices and lasted between 45 and 60 minutes. All but one practice chose to have lunchtime meetings. Each practice decided which members of the practice team would be present. The number of participants ranged from two to eleven and always included a health visitor and GP. Other team members present included practice nurses, practice managers, clerical staff and, in two practices, a midwife.
We succeeded in meeting with all the action research practices. From information provided by the research team and through facilitated discussion, the primary health care staff identified aspects of their provision of child health surveillance which they wished to improve. A mean of two decisions per meeting were made and at only two of 42 meetings was no new change to child health surveillance agreed. Examples of the decisions that were made are shown in Table 2. According to the practice teams, the majority of these decisions were implemented during the study period.
|
Ten action research practices decided to audit their coverage of CHS reviews as a result of the written feedback that was provided by the research team. Most of these practices found that data collected at a practice level showed a higher coverage of CHS reviews than that collected at health authority level by the research team. Discrepancies were due mainly to failure of the paper copies of CHS reviews reaching the central collection point, but mobility of practice populations was also a factor.
The primary health care teams said that they valued feedback that was accurate and anonymous, but allowed comparison of practice performance. They wanted a minimum of written information and preferred graphs or tables. Quarterly feedback reports were of little use to practices with low birth rates, e.g. one or two births per quarter, and for these practices annual feedback would have been preferred.
Through discussion with primary care staff in the study, a standard for child health surveillance was developed. This included measures of structure, process and outcome. The teams thought the completed standard was acceptable and achievable. Discussion with the practice teams also generated many good ideas on "how to run baby clinics", "how to organize child health surveillance" and "promoting health education for the under fives". These suggestions were collated and made available to all practices in the health authority.
Interviews
A health visitor and GP from each of the 28 study practices were interviewed before and after the provision of feedback to the practices. On the second occasion, in five practices a different GP was interviewed due to changing responsibilities for child health surveillance within the practice, and in six practices a different health visitor was interviewed because the first health visitor had left.
In the second interviews, more health visitors and GPs in the action research practices than in the control practices reported changes having been made to CHS organization, audit, communication and use of the PCHR (Table 3). More health visitors and GPs in the control practices than in the action research practices said that there were communication difficulties between them, (6 versus 0 practices; chi-square = 4.67, P = 0.02).
|
Ten health visitors and 10 GPs in the 14 action research practices thought that being involved in the research process had been beneficial. The health visitors felt empowered, and both health visitors and GPs thought that the profile of child health surveillance had been raised in the practices. Only one health visitor and GP said that written feedback on practice performance would have been sufficient and preferable to the facilitated practice meetings. Most of the health visitors and GPs thought that the combination of written feedback on practice performance and facilitated practice meetings had been very useful. In comparison, only four health visitors and four GPs in the control practices thought that written feedback alone had been useful. Examples of positive and negative comments about involvement in the study are shown in Tables 4 and 5
. More positive comments were made by health visitors and GPs in the action research arm of the study.
|
|
Baby clinics
There were no observable improvements in the structure or process of the practice baby clinics either when all practices were compared before and after feedback, or when the action research practices were compared with the control practices.
Parent questionnaires
Questionnaires were sent to all 534 parents of babies aged between 3 and 6 months registered with the study practices immediately before the year of feedback began; 401(75%) were returned. After the feedback year, questionnaires were again sent to all parents with babies aged 3–6 months registered with the study practices, i.e. a different set of parents. Of 498 questionnaires sent, 367 (74%) were returned.
Responses of parents from the action research practices were compared with those of parents from the control practices. Having corrected for cluster,27,28 there were no significant differences after feedback that had not been present before feedback.
Return rates of child health surveillance reviews
There are four recommended CHS reviews conducted during the first year of a child's life: a birth review completed by the midwife, 10-day review completed by the health visitor, 6–8-week review completed by the GP and 6–9-month review completed by the health visitor.30 These reviews were considered separately during data analysis.
During the feedback year, 2015 babies were born and registered to the 28 study practices. The mean birth rate per practice was 75 (range 14–200). Considering the total study sample, the return rates of the four different reviews did not improve significantly during the feedback year. The return rates of CHS reviews per practice before and after the year of practice feedback were analysed using Poisson regression. There was no association between the return rate of each of the four child health surveillance reviews and being an action research practice.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
If the data generated from the practice meetings and the interviews with health visitors and GPs are considered, using an action research model to involve primary health care staff in the research process was successful. It promoted audit and initiated changes in the provision of services within the practices. Involvement in the action research process was associated with better communication between health visitors and GPs. The majority of health visitors and GPs thought it a useful process and valued having an opportunity for their views to be heard.
Our action research model met most of the criteria in the definition of action research described in the introduction to this paper. Practitioners were actively in-volved in the research process. They decided which team members would participate and what issues they wanted to address. In facilitating the practice meetings, we focused on change and improvement without being directive. Questions arose either directly from practitioner's experience or from the cycle of feedback data provided by the research team. Although we collected most of the data, the majority of practices were prompted to audit child health surveillance themselves as a result of feedback.
In terms of the spectrum of action research methodology, our study was more ‘experimental’ than ‘empowering’,14 or more ‘technical’ than ‘emancipatory’.15 As researchers, we designed the project and pre-determined its timetable and outcome measures. The project would have been more collaborative if we had involved the participants in its design. Traditional or more positivist models10 of action research are described as ‘experimental/technical’, with external researchers exercising greater control. ‘Empowering’ or ‘emancipatory’ action research is more collaborative, with emphasis on the practitioner as researcher. The methodology is more in line with the critical social science research paradigm.14,15
Although our study was at the ‘experimental/technical’ end of the action research spectrum, the health visitors and GPs involved thought the process had been useful in enabling them to reflect on and change their provision of child health surveillance. There are many examples from outside medicine of action research being successful in promoting change and bridging the gap between research and practice.10 –17 Action research fits well with the principles of adult learning theory, e.g. participation, reflection on practice, use of participants' own experience and being problem-based.31 Both in education and nursing there is a current emphasis on reflective practice, and action research can be seen as a natural extension of this.32 Involvement of practitioners leads to a sense of ownership, and the group process provides motivation for change.10,15,16,33 Having compared models of organizational change with action research, Zuber-Skerritt concludes that emancipatory action research is organizational change ‘best practice’.15
Using feedback which enables comparison with other practices should also motivate change.2,10,33 However, in our study, feedback on a practice's own performance produced most decisions for change, with a practice's own patients' views generating the most interest.
In the medical literature on feedback interventions, little emphasis has been given to the type of feedback preferred by health professionals. We found that brief written feedback using graphs and tables rather than words was most acceptable. The practices wanted to be able to compare their performance with that of other practices whilst maintaining anonymity. Meetings to discuss written feedback were thought to be more valuable than written feedback alone, which contradicts one previous study of the effectiveness of feedback strategies in general practice.34
As a research team, we found that involving the practices provided us with greater understanding of the practical issues in providing child health surveillance in primary care. It enabled us to develop feedback that was useful to the practices and allowed us to produce a standard for child health surveillance that was seen as acceptable and achievable by practice teams. It has since been incorporated in the Nottingham Health Authority guidelines for child health promotion.35
Practical difficulties in our use of action research to promote change in primary care included the collection of accurate data on practice performance, the organization of practice meetings and staff turnover. From their own audits, the practices found that centrally collected data were incomplete and were less accurate in practices with a high patient turnover. The turnover of practice staff was surprisingly high during the study period. This must influence effective teamwork and the quality of service provision in the practices, as well as data collection and the influence of feedback. Meyer also found that an ever changing workforce made effective collaboration in action research and negotiation of change very difficult.17
From our outcome measures of change, i.e. the baby clinic observations, parent questionnaires and return rates of CHS reviews, it could be concluded that using action research was not successful. The difficulty of measuring the effectiveness of interventions to improve quality of care is well recognized.2,3,5,34 In our study, initial standards were high, allowing little room for improvement. The small size of the study meant insufficient power to detect all but large effects as statistically significant and the adjustment for cluster reduced the power of the parent questionnaire survey. Although there was a high level of self-reported change in the practices, the changes made were usually different in each practice, which could also account for difficulty in demonstrating that change had occurred in the group as a whole.
There continues to be a need for research on improving quality of care in health and both qualitative and quantitative techniques will be required.2,3,5,8 From experience in other fields and using the example of our study, action research can be a useful method for investigating and encouraging improvement in health care. We conclude that using action research to involve primary care teams in the evaluation of service provision promotes teamwork, decision making, audit and change.
|
Acknowledgments |
|---|
The study was funded by The Department of Health. We would like to thank Nottingham Community Health NHS Trust for their permission to do the research, the study practices, Edna Gibson and Ann Zamorski for their data entry, and staff from the Information Service Department of Nottingham Community Health NHS Trust and the Department of Information Management and Technology at Queen's Medical Centre. We also valued Lindsay Groom's comments on earlier drafts of this paper.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
1 NHS Executive. The new NHS: a national framework for assessing performance. Leeds: NHSE, 1998.
2
Donabedian A. The effectiveness of quality assurance. Int J Qual Health Care 1996; 8: 401–407.
3
Grol R. Research and development in quality of care; establishing the research agenda. Qual Health Care 1996; 5: 235–242.
4
Mulrow C, Cook D, Davidoff F. How much are we willing to pay for information about quality of care? Ann Internal Med 1997; 126: 391–393.
5 Oxman A, Thompson MA, Davis DA, Haynes RB. No magic bullets: a systematic review of 102 trials of interventions to help health care professionals deliver services more effectively or efficiently. Can Med Assoc J 1995; 153: 1423–1431.[Abstract]
6 Haines A. The science of perpetual change. Br J Gen Pract 1996; 46: 115–119.[Web of Science][Medline]
7 Grol R, Baker R, Roberts R, Booth B. Systems for quality improvement in general practice. Eur J Gen Pract 1997; 3: 65–68.
8 Wensing M, Van Der Weijden T, Grol R. Implementing guidelines and innovations in general practice: which interventions are effective? Br J Gen Pract 1998; 48: 991–997.[Web of Science][Medline]
9
Grol R. Beliefs and evidence in changing clinical practice. Br Med J 1997; 315: 418–421.
10 Lewin GW (ed.). Resolving Social Conflicts; Selected Papers on Group Dynamics by Kurt Lewin. New York: Harper and Brothers, 1948.
11 Carr W, Kemmis S. Becoming Critical: Education, Knowledge and Action Research. London: Falmer Press, 1986.
12 Whyte WF. Participatory Action Research. New York: Sage, 1991.
13 Elliott J. Action Research for Educational Change. Buckingham: Open University Press, 1991.
14 Hart E, Bond M. Action Research for Health and Social Care: a Guide to Practice. Buckingham: Open University Press, 1995.
15 Zuber-Skerritt O. New Directions in Action Research. London: Falmer Press, 1996.
16 Titchen A, Binnie A. What am I meant to be doing? Putting practice into theory and back again in new nursing roles. J Adv Nurs 1993; 18: 1054–1065.[Web of Science][Medline]
17 Meyer JE. New paradigm research in practice: the trials and tribulations of action research. J Adv Nurs 1993; 18: 1066–1072.[Web of Science][Medline]
18
Murray SA, Tapson J, Turnbull L, McCallum J, Little A. Listening to local voices: adapting rapid appraisal to assess health and social needs in general practice. Br Med J 1994; 308: 698–700.
19 Robinson L, Stacy R. Palliative care in the community: setting practice guidelines for primary care teams. Br J Gen Pract 1994; 44: 461–464.[Web of Science][Medline]
20 Department of Health and the Welsh Office. General Practice in the National Health Service. The 1990 Contract. London: HMSO, 1989.
21 Polnay L, Pringle M. General practitioner training in paediatrics in the Trent region. Br Med J 1989; 298: 1434–1436.
22 Marsh G, Russell D, Russell I. Is paediatrics safe in general practitioners' hands? A study in the North of England. J R Coll Gen Pract 1989; 39: 138–141.[Web of Science][Medline]
23 Butler J, Freidenfeld K, Relton J. Child Health Surveillance in the New NHS. University of Kent at Canterbury: Centre for Health Service Studies, 1995.
24 Brazier S, Hammond R, Waterman SR. A New Geography of Nottingham. Nottingham: Trent Polytechnic, 1984.
25 Glaser BG, Strauss AL. The Discovery of Grounded Theory. Chicago: Aldine, 1967.
26 Lincoln Y, Guba E. Naturalistic Inquiry. California: Sage, 1985.
27
Donner A, Birkett N, Buck C. Randomization by cluster: sample requirements and analysis. Am J Epidemiol 1981; 114: 906–914.
28 Hauck W, Gilliss C, Donner A, Gortner S. Randomization by cluster. Nurs Res 1991; 40: 356–358.[Web of Science][Medline]
29 Jarman B. Identification of underprivileged areas. Br Med J 1983; 286: 1705–1709.
30 Hall DMB. Health for All Children. 3rd edn. Oxford: Oxford University Press, 1996.
31 Brookfield SD. Understanding and Facilitating Adult Learning. Milton Keynes: Open University Press, 1986.
32 Schön DA. The Reflective Practitioner. New York: Basic Books, 1983.
33 Hertzberg F. Work and the Nature of Man. New York: World Publishing, 1966.
34 Szczepura A, Wilmot J, Davies C, Fletcher J. Effectiveness and cost of different strategies for information feedback in general practice. Br J Gen Pract 1994; 43: 19–24.
35 Nottingham Health Authority and Nottingham Community Health NHS Trust. Nottingham Programme for Child Health Promotion 0–19 Years. Nottingham: Nottingham Community Health NHS Trust, 1997.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. I. Loureiro, A. R. Goes, G. Paim da Camara, M. Goncalves-Pereira, T. Maia, and L. Saboga Nunes Priorities for mental health promotion during pregnancy and infancy in primary health care Global Health Promotion, March 1, 2009; 16(1): 29 - 38. [Abstract] [PDF]
|
|
|
|
 |
|
 T. J. Mendenhall and W. J. Doherty Partners in Diabetes: Action research in a primary care setting Action Research, December 1, 2007; 5(4): 378 - 406. [Abstract] [PDF]
|
|
|
|
 |
|
 E C Goyder, L Blank, E Ellis, A Furber, J Peters, K Sartain, and C Massey Reducing inequalities in access to health care: developing a toolkit through action research Qual. Saf. Health Care, October 1, 2005; 14(5): 336 - 339. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Whitehead, A. Taket, and P. Smith Action research in health promotion Health Education Journal, January 1, 2003; 62(1): 5 - 22. [Abstract] [PDF]
|
|
|
|
 |
|
 M. H. Bellman Commentary Adv. Psychiatr. Treat., January 1, 2001; 7(1): 40 - 42. [Full Text]
|
|
|
|
 |
|
 M. BLAIR Public health: The need for and the role of a coordinator in child health surveillance/promotion Arch. Dis. Child., January 1, 2001; 84(1): 1 - 5. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||