Family Practice Advance Access originally published online on May 25, 2008
Family Practice 2008 25(3):162-167; doi:10.1093/fampra/cmn020
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
What may help or hinder the implementation of computerized decision support systems (CDSSs): a focus group study with physicians
a Department of General Practice and Primary Care, University of Helsinki, Helsinki
b University of Tampere
c FinOHTA and Paediatric Research Centre, Tampere University Hospital and University of Tampere
Correspondence to Helena Varonen, Department of General Practice and Primary Care, University of Helsinki, Kytösuontie 11, PL 49 00014 Helsinki, Finland; Email: helena.varonen{at}helsinki.fi
Received 4 October 2007; Revised 18 March 2008; Accepted 14 April 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
Objectives. To identify potential barriers and facilitators to implementing computerized decision support systems (CDSSs) in health care as perceived by clinicians.
Methods. We carried out a qualitative focus group study with primary and secondary health care settings in six areas of Finland. A total of 39 interviewed physicians, of whom 22 practised in primary care and 17 in secondary care. The main outcome measures physicians’ expectations, preconceived barriers and facilitators were explicitly identified by the participants during the interviews.
Results. Identified barriers were earlier experience of dysfunctional computer systems in health care, potential harm to doctor–patient relationship, obscured responsibilities, threats to clinician's autonomy and potential extra workload due to excessive reminders. Identified facilitators were self-control of frequency and contents of CDSS and noticeable help of CDSS in clinical practice. It was easy for the physicians to think of applications and clinical topics for CDSS that could help them to avoid mistakes and improve work processes.
Conclusions. Physicians had relatively positive attitudes towards the idea of CDSS. They expected flexibility, individuality and reliability of the CDSS. The rather high level of computerized practices and wide use of electronic guidelines probably have paved the way for the CDSS in Finland.
Keywords. Decision support, computerized, focus group, barriers, facilitators.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
Computerized decision support systems (CDSSs) are developed to provide patient-specific, evidence-based advice for practising clinicians. It is hoped that many current challenges in health care could be overcome with CDSS, including the information overload, practice improvement,1 avoidance of treatment errors and costs.2 For a clinician, the CDSS may serve as a data management tool and a memory aid. The effectiveness of CDSS has been studied in two systematic reviews,3,4 and there is solid evidence that CDSS can significantly improve practice performance in health care. The efficacy of CDSS has been shown in many specific clinical problems. However, implementation of a generic, multi-purpose CDSS, which covers several disease entities and is evidence based and applicable in both primary and specialized care, is in its early phase in most health care systems.
Current electronic patient records lack opportunities to retrieve the information buried in them at the point-of-care. In drug therapy, all background information, both patient specific and that of best practices, is often inaccessible to the physician, and consequently, optimal therapy may not be chosen.5 A just-in-time CDSS may link patient-specific observations recorded in a structured way with clinical guidelines and bring reminders, alerts or prompts to the physician at the moment when the information is needed. In Finland, the Evidence-Based Medicine electronic Decision Support (EBMeDS) project was set up to develop and study a CDSS that is generic, frequently updated and a national web-based service.
The EBMeDS project has its knowledge base in the large collection of Finnish practice guidelines. Decision support is developed by the two guidelines groups at the Finnish Medical Society Duodecim. The government funds the Current Care guideline programme (http://www.kaypahoito.fi/) producing evidence-based national guidelines. There were 73 comprehensive Current Care guidelines in the beginning of 2007. EBM Guidelines (http://ebmg.wiley.com) is a collection of 1200 concise and frequently updated guidelines especially for primary care.6 Although these guidelines are much used e.g. via the Internet, clinicians may not always recognize new evidence and situations where practice should be modified. Therefore, active guideline implementation methods such as reminders are needed.7
To be able to introduce CDSS in health care, we need to understand users’ perspectives and preferences on the new information technology (IT). According to previous studies, user resistance is a major obstacle in the adoption of CDSSs.8–10 To include user perspectives11,12 in planning a generic just-in-time CDSS in the EBMeDS project, we carried out a focus group study among the future end-users. In this paper, we report the barriers and facilitators that the Finnish physicians conceived upon the implementation of CDSS.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
The study was carried out between October and December 2005. We decided on qualitative focus group method because it can qualify the participants’ responses and identify important contingencies associated with their answers better than quantitative methods.13,14 We organized four focus groups in larger towns with university clinics, two in a small town and one in a rural health centre adjacent to a larger town. Very small rural units were not represented in the sample.
Participants
We interviewed a total of 39 physicians in 7 focus groups. We used our contacts and those of the network of the Centre for Pharmacotherapy Development (http://www.rohto.fi/) to form the groups. These contact physicians were asked to invite four to eight clinicians working in the area to the interview. They were to include physicians with different backgrounds, representing different viewpoints and willing to exchange their views.
Interviews
We had a semi-structured guide that was followed in each interview (Box 1). The interview schedule was developed after the initial literature searches and discussions within the EBMeDS study group. All writers are members of the EBMeDS study group. HV was and is one of the editors of the reminders of the CDSS, and MK is the chair of the Advisory Committee. Interviews were made flexible to allow participants contribute according to their role and experience. MK moderated all interviews. HV was the facilitator, who posed additional questions and supported the more quiet participants and made field notes of the conversations. The focus group interviews lasted from 90 to 120 minutes.
BOX 1—Interview schedule.
Starting off
What topics do you wish to have decision support for?
What are the topics where you do not want decision support? What are the potential advantages of decision support? What are the potential disadvantages or harms of decision support? How often do you think you could need/use/tolerate decision support? What factors facilitate using decision support? What factors prevent using decision support?
|
Data analysis
The interviews were audiotaped and transcribed verbatim. The transcripts and the field notes were read independently by all three of us, and the following steps were also done in triplicate with iterative discussions. First, we coded the themes in each interview independently. Then, we discussed the initial themes and range of responses in order to produce an initial framework for analysis.13–15 We grouped emerging themes by category. The information in each category was condensed, reflected and interpreted together by all researchers. The basic idea was to find all relevant factors that the individual physicians brought up in the discussions and that were further explored by the others. In the analysis, we interpreted the items, which were discussed as potential facilitators or barriers to CDSS. We also interpreted the discussions comparing results in primary and secondary care.
Approval
The study was approved by the local health care managers, and it was conducted according to local ethical regulations. All participating physicians gave their written informed consent.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
Characteristics of the informant physicians appear from Table 1. There were only primary care physicians in four groups, only secondary care physicians in one group and two groups were mixed.
|
Use of CDSS
After a short introduction to CDSS, interviews were started with an open question on potential usage for CDSS (Box 1). Physicians in all focus groups could easily produce ideas of situations where CDSS might be helpful, either in clinical or administrative issues. ‘Allergies, of course, but also risk of thrombosis, and other risks associated to anesthesia ...’ (male, secondary care). Examples of applications suggested by the interviewees are gathered in Table 2. Secondary care physician had a specific interest in topics that produce alerts to prevent potentially life-threatening situations, e.g. thrombosis prevention and drug-interactions alerts. Primary care physicians suggested applications to monitor chronic disorders, diabetes, hypertension etc. ‘It would be good to have reminders of things that you need to check yearly in diabetic patients. You may think you just checked the reflexes and so on but it may be a long ago ...’ (female, primary care). Both physician groups were more willing to accept reminders in diseases with a solid and specific knowledge base than in rare conditions and in those where the management is rapidly being evolved.
|
In all focus groups, administrative topics for the use of CDSS were brought up. Recurring themes were helpful for drawing up certificates, reminders of secondary care referral criteria, monitoring of care queues, reminders of checklists for hospital discharge and of the existence of documents such as living will (Table 2).
Opinions on CDSS as a tool in diagnostic processes were divided. There was concern of excessive reminders of rare disorders: ‘If it starts to warn you of some rare disorders all the time, that will cause anxiety to the physician and unnecessary lab tests, too ...’ (male, primary care). ‘My guess is that the biggest problem is that if the computer gives you a suggestion of diagnosis, as a young physician, you need to test, you cannot discard ...’ (female, primary care). That CDSS could lead to unnecessary diagnostic testing was repeatedly brought up, but physician-initiated passive CDSS in difficult diagnostic cases was considered a usable opportunity. Both primary and secondary care physicians had grave concerns of reminders on specific diagnostic issues. ‘The programme cannot be such that it suggests all kinds of nonsense all the time, no, the topics need to be essential ...’ (male, primary care).
Advantages and disadvantages
As the greatest advantage of CDSS, the physicians looked forward to help in managing the mass and complexity of clinical information embedded in the electronic patient record. ‘With the amount of information to consider in clinical work, however, human brain as clever as it can be is limited ... CDSS could organise and manage the information so that it could be retrieved when needed’ (male, secondary care). Physicians also thought that CDSS could help to prevent overlapping work such as inquiring and registering the same data from one patient several times. Physicians thought that repeated reminders could serve educational purposes, enhance developing work processes and help in harmonizing the varying practices. ‘As a matter of fact, it [CDSS] teaches you. If you get a reminder saying that you cannot prescribe this neurolept to a demented patient, you'll learn it. My opinion is that it improves clinical skills, not ruins them’ (female, secondary care).
The primary care physicians thought that getting CDSS reminders before the consultation would be helpful. For example, CDSS could be used to check laboratory findings, run drug-interactions alerts and give suggestions before the patient encounter. ‘It would be practical to run the check ups during the night before the visit and have a listing of alerts and recommendations with the patient list. Then you could be better prepared to see the patient’ (male, primary care).
The physicians were aware of numerous potential threats of CDSS. One recurring disadvantage was the possible harm to doctor–patient relationship and clinical skills. ‘Does it mean that you are even more closely tied with the computer, and the patient sits quietly behind your back? There is also risk that you stop using your ‘clinical eye’, touching and listening to the patient’ (male, secondary care). There was concern of the potential extra workload; a large number of reminders may lengthen the days and add to the already massive amount of work. ‘There must be a balance, enough reminders that you can benefit of and not too much extra work and checking’ (female, primary care).
All focus groups discussed the effects of CDSS on safety and responsibility issues. With erroneous input, e.g. out of date drug lists in the patient record system or malfunctioning software, the reminders may be false and cause harm. External factors, such as computer viruses, may threaten the system. Mechanical decision making and too much trust on the system either by physicians or by patients may diminish independent thinking and decision making. ‘What scares me is that you start to trust the computer too much, and your clinical skills start to deteriorate or do not develop at all ...’ (male, primary care).
Barriers and facilitators
The identified barriers and facilitators of implementing CDSS are presented in Table 3. The three most often recurring barriers to CDSS implementation included earlier experience of dysfunctional computer systems in health care, general resistance towards changes in practice and time management issues. ‘All that is new, sounds like extra work ...’ (female, primary care). Physicians were concerned of the data registering needs and potential excessive secretarial work. ‘It may be that the secretarial work is allocated to physicians, but I think that physician should do physician's job, whatever it is then ...’ (male, secondary care). All groups listed numerous examples of dysfunctional health care IT systems, and they were worried of consequences if CDSS is built on the existing imperfect systems. ‘It may be problematic if the current system does not function adequately and then CDSS is added on ... We would need one standard, one functional programme that would serve adequately ...’ (male, primary care).
|
As facilitators for CDSS the physicians brought up the following factors: that users and trusted peers develop and produce CDSS; that reminders are clear and concise; and that interfaces are easily learnt and user-friendly (Table 3). ‘Surely, what we would want would be a user friendly system where you have the information you need with just a few clicks ...’ (female, secondary care) ‘The reminders should be behind one button, visible if you want them ...’ (female, primary care) ‘What is the most important thing is that the system is simple, not complex as the current ones that are developed by engineers ...’ (male, primary care) ‘This is the spot for education, so that we know what we treat and register things the same way ...’ (male, primary care).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
Finnish physicians interviewed in focus groups appeared to be ready to accept the implementation of generic CDSS provided that they have some control over the system. They expected flexibility, individuality and reliability of the CDSS. They could easily produce suggestions for CDSS applications to improve their work processes and worries based on their past IT experiences. In addition to threats to clinician's autonomy and potential harm to doctor–patient relationship, the earlier experience of dysfunctional computer systems was the strongest perceived barrier to the effective implementation of CDSS.
Generally, the interviewed physicians were interested in the planned generic CDSS and rich in examples where it could facilitate their daily practice. The perceived attitudes were more positive than those presented in the study of Rousseau et al.8 and in that of Zheng et al.10 The physicians eagerly expressed needs, clinical situations and work process issues that could be managed with better functioning IT and CDSS. The younger physicians had the most enthusiastic attitudes, and the criticisms were mostly presented by older physicians and those from primary care. This latter is probably related to the fact that primary care physicians have longer practical experience with computerized patient records than those working in secondary care.
Physicians’ main worries of CDSS concerned harms to doctor–patient relationship in the form of effects on communication in the consultation room and problems with erroneous data input and malfunctioning systems. Studies of effects of IT use on doctor–patient communication are surprisingly scant. However, in the study by Hsu et al.,16 computers had mainly positive effects on physician–patient interactions in the form of improved communication of medical care. Again, the worries of deleterious effects on communication may become of lesser importance as the digital natives take over, as patients as well as professionals.
Avery et al.17 studied ways to improve general practice computer systems. As priorities, they identified improving the knowledge base for clinical decision support, paying greater attention to human ergonomics in system design and focus on accurate data recording. Trusted knowledge base is considered one key issue in effective CDSS.18,19 The interviewed physicians were active users of electronic evidence-based guidelines. Logbook data indicate that guidelines are consulted in Finland more frequently than anywhere in the world.6 To offset the potential harms of CDSS, a trusted knowledge base is only one component; special focus needs to be put on user-friendliness. Correct data input has to be the responsibility of the professionals, who should be motivated to avoid potential harms of erroneous reminders. A structured electronic patient record with nationally agreed core items is on the way in Finland and will help professionals in this. On the other hand, structuring means a lot of clicking and can prevent achieving these advantages.
The strengths of this qualitative study include that we interviewed both primary and secondary care physicians in different parts of our country. The interviews were intense and deep, and the same framework was utilized in all of them. A possible weakness is that we only interviewed physicians, and furthermore computer or technology enthusiasts may have been over represented. Nurses and other health care professionals’ viewpoints are as important as those of physicians in the implementation of CDSS. However, when designing these first focus groups, it was more feasible to aim at homogeneity in sampling between groups and as much heterogeneity within the groups as possible.13
Another weakness in this study is that the physicians had only experience of CDSS in the form of electronic guidelines and drug-interaction alerts in patient record systems, and assumptions of the benefits and harms of a just-in-time, patient-specific CDSS were based on prospected thinking only. The facilitators may have had an influence on physicians’ attitudes. Both have a known history in guidelines production and are involved in the EBMeDS project. This may have restricted expression of the most critical comments. However, a lot of the criticisms stemmed from experience with the present electronic systems, with which there are no conflicts of interest.
The findings of this focus group study will be used in the development of the generic, multi-purpose CDSS in the EBMeDS project. It may be good that the interviewers are deeply involved with the project, ensuring the application of these results in the development process. Key conclusions for active implementation include that the CDSSs need to be planned flexible; physicians expect prompts that are severity graded and systems that allow turning off certain types of reminders. The selection of clinical applications proposed by interviewed physicians will be used in prioritizing the CDSS topics. We intend to continue using and updating user perspectives in the project and in the integration of CDSS with various patient record systems.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
In conclusion, the Finnish physicians interviewed in this qualitative study had rather positive attitudes towards implementation of CDSS provided that they have some control over the system. They expected flexibility, individual tailoring and reliability of the CDSS. The rather high level of computerized practices and wide use of electronic guidelines probably have paved the way for the CDSS in Finland.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
Funding: The study was partially funded by the Centre for Pharmacotherapy Development Rohto in Finland.
Ethical approval: None.
Conflicts of interest: None.
|
Acknowledgments |
|---|
We thank all the physicians that participated in and helped to organize the interviews. The comments and support from the EBMeDS study group have been invaluable.
|
Notes |
|---|
Varonen H, Kortteisto T and Kaila M. What may help or hinder the implementation of computerized decision support systems (CDSSs): a focus group study with physicians. Family Practice 2008; 25: 162–167.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsDeclarationReferences |
|---|
1 Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision support systems on physician performance and patient outcomes: a systematic review. J Am Med Assoc (1998) 280:1339–1346.
2 McMullin ST, Lonergan TP, Rynearson CS, Doerr TD, Veregge PA, Scanlan ES. Impact of an evidence-based computerized decision support system on primary care prescription costs. Ann Fam Med (2004) 2:494–498.
3 Garg AX, Adhikari NK, McDonald H, et al. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: a systematic review. J Am Med Assoc (2005) 293:1223–1238.
4 Kawamoto K, Houlihan CA, Balas EA, Lobach DF. Improving clinical practice using clinical decision support systems: a systematic review of trials to identify features critical to success. Br Med J (2005) 330:765.
5 Bennett JW, Glasziou PP. Computerised reminders and feedback in medication management: a systematic review of randomised controlled trials. Med J Aust (2003) 178:217–222.[Web of Science][Medline]
6 Varonen H, Jousimaa J, Helin-Salmivaara A, Kunnamo I. Electronic primary care guidelines with links to Cochrane reviews—EBM Guidelines. Fam Pract (2005) 22:465–469.
7 Grimshaw JM, Thomas RE, MacLennan G, et al. Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess (2004) 8(6):iii–iv. 1–72.[Medline]
8 Rousseau N, McColl E, Newton J, Grimshaw J, Eccles M. Practice based, longitudinal, qualitative interview study of computerised evidence based guidelines in primary care. Br Med J (2003) 326:314.
9 Short D, Frischer M, Bashford J. Barriers to the adoption of computerised decision support systems in general practice consultations: a qualitative study of GPs’ perspectives. Int J Med Inform (2004) 73:357–362.[Web of Science][Medline]
10 Zheng K, Padman R, Johnson MP, Diamond HS. Understanding technology adoption in clinical care: clinician adoption behavior of a point-of-care reminder system. Int J Med Inform (2005) 74:535–543.[CrossRef][Web of Science][Medline]
11 Carroll C, Marsden P, Soden P, Naylor E, New J, Dornan T. Involving users in the design and usability evaluation of a clinical decision support system. Comput Methods Programs Biomed (2002) 69(2):123–135.[CrossRef][Web of Science][Medline]
12 Ahearn MD, Kerr SJ. General practitioners’ perceptions of the pharmaceutical decision-support tools in their prescribing software. Med J Aust (2003) 179(1):34–37.[Web of Science][Medline]
13 Barbour RS. Using focus groups in general practice research. Fam Pract (1995) 12:328–334.
14 Barbour RS. The role of qualitative research in broadening the "evidence base" for clinical practice. J Eval Clin Pract (2000) 6(2):155–163.[CrossRef][Web of Science][Medline]
15 Pope C, Ziebland S, Mays N. Qualitative research in health care. Analysing qualitative data. Br Med J (2000) 320:114–116.
16 Hsu J, Huang J, Fung V, Robertson N, Jimison H, Frankel R. Health information technology and physician-patient interactions: impact of computers on communication during outpatient primary care visits. J Am Med Inform Assoc (2005) 12:474–480.[CrossRef][Web of Science][Medline]
17 Avery AJ, Savelyich BS, Sheikh A, Morris CJ, Bowler I, Teasdale S. Improving general practice computer systems for patient safety: qualitative study of key stakeholders. Qual Saf Health Care (2007) 16(1):28–33.
18 Bates DW, Kuperman GJ, Wang S, et al. Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality. J Am Med Inform Assoc (2003) 10:523–530.[CrossRef][Web of Science][Medline]
19 Purcell GP. What makes a good clinical decision support system. Br Med J (2005) 330:740–741.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Wright, D. F Sittig, J. S Ash, S. Sharma, J. E Pang, and B. Middleton Clinical Decision Support Capabilities of Commercially-available Clinical Information Systems JAMIA, September 1, 2009; 16(5): 637 - 644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Delaney Evidence-based diagnosis in general practice: needs both robust evidence and sophisticated electronic health record systems Fam. Pract., August 1, 2009; 26(4): 239 - 240. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||