The use of information technology (IT) in the UK's National Health Service (NHS) has a long history. Since the publication of the Information for Health Strategy1 and more recently England's ‘National Programme for IT’ (NPfIT), computer applications within the NHS have been increasing.2 In Wales related developments have been undertaken by Informing Healthcare3 and in Scotland by eHealth for NHS Scotland.4 These developments are not limited to the UK but have international significance. Around the world similar developments in IT use are taking place with increased employment of electronic health records (EHRs), automated administration and increased electronic sharing of patient information. The importance of IT in the NHS modernization agenda has been underpinned by several reports (such as those led by Derek Wanless in 20025 and 20076) and its role in increasing patient safety and reducing errors made clear.7
There is also a large and growing body of literature about aspects of health informatics related to policy, hardware, software and implementation. One of the factors identified as significant in the introduction of information technology into health care practice is the attitude of staff that will be required to use it. In the UK surveys undertaken by Medix (a market research company in the health sector) found that the attitudes of doctors and nurses have shown increasingly negative attitudes.8,9
The Technology Acceptance Model (TAM)10,11 may provide a context to study the acceptance, or otherwise, of IT in the NHS. It has also been extended by Dixon and Dixon into the Information Technology Adoption Model (ITAM),12,13 which may provide a more comprehensive theoretical model for this study.
Over twenty years ago, Stronge and Brodt14 were studying this area in the USA with their Nurses’ Attitudes Towards Computers (NATC) questionnaire and others have continued this work, using similar instruments with different findings. Sultana15 and McBride and Nagle16 found more positive attitudes than Stronge and Brodt.14 Schumacher et al.17 found that students showed more positive attitudes than qualified staff, but Scarpa and Smeltzer18 found no differences in attitude with nursing experience or educational level, but had found that experience with computers was significant. In addition, Schwirian et al.19 had shown gender differences in attitude.
A recent review of the literature on eHealth has been carried out,20 but little was included on the attitudes of staff and students towards these emerging technologies.
In developing a search strategy, bias may arise from a number of factors, including the predilection of journals to publish research with (statistically) significant findings; for example, medline bias for excluding papers not indexed in medline and ‘language bias’ from exclusion of non-English publications.21 In our review, the selection of appropriate sources and the development of the search strategy was a dynamic, iterative process balancing the expertise and neutrality of the librarian with the knowledge and experience of the researchers. The key principle in our selection was to be as inclusive as possible within the resources and timescales for the project whilst being open to the possibility of bias. In developing our search strategy, we continually traded sensitivity (recall) against specificity (precision).
Thirteen databases (see Table 1) were identified as covering relevant material for this review. These databases were explored in a series of discussions between the researchers and librarian, which balanced the knowledge of the librarian on their characteristics (for example, accessibility, currency and potential relevance) and the researchers’ knowledge of the subject area. Potential search terms were applied across all these databases, enabling us to explore the suitability and functionality of their search interface and the significance of their output to our review.
Table 1. Databases included in the review
|Allied and Complementary Medicine Database (amed)*|
|British Nursing Index (BNI)*|
|Cumulative Index to Nursing and Allied Health Literature (cinahl)*|
|Health Management Information Consortium (HMIC) *|
|Index to theses|
|Maternity and Infant Care (MIDIRS)*|
|NHS National Research Register|
|ZETOC conference search|
Developing the search strategy
In tandem with our exploration of databases, a cyclic, iterative approach to identifying appropriate search terms was undertaken. Search terms were identified through discussion between the research team, including the faculty librarian, and by scanning background material. The search strategy is detailed in Table 2.
Table 2. Search strategy for database searches
|information tech*||nurs*||attitud*||research*||patient attitud*|
|communication tech*||doctor*|| ||study||patients attitud*|
|ICT||physician*|| ||studies||patient information*|
|comput*||professional*|| ||evaluat*||consumer attitud*|
|e health||practitioner*|| || ||consumers attitud*|
|Ehealth||clinician*|| || ||consumer information*|
|Informatics||staff|| || ||public attitud*|
|e learn*||health care worker*|| || ||publics attitud*|
|elearn*||personnel|| || ||public information*|
|technology in health care||therap*|| || || |
|midwi*|| || || |
|dietician*|| || || |
|dietitian*|| || || |
|nutritionist*|| || || |
|radiograph*|| || || |
|radiotherap*|| || || |
|radiolog*|| || || |
|surgeon*|| || || |
|scientist*|| || || |
|health visitor*|| || || |
|dentist*|| || || |
|dental|| || || |
|physiotherap*|| || || |
|GP|| || || |
|medic*|| || || |
Search terms were developed using what were considered to be the richest sources of data for this review; amed, BNI, cinhal, embase, medline, Sportdiscus and PsycInfo (‘core databases’ accessed on the Ovid interface).
The initial search was built on four concepts that were to be present in all retrieved references as they were seen as being key to the research questions:
- 2health professionals;
The concepts were combined together with AND in the search strategy.
On exploring the references retrieved using the four initial concepts, a fifth concept was added—Public/Patient attitudes. This was combined with a NOT in order to exclude items that related to the public's attitude to IT in health care.
Overall, a sensitive approach was taken to the development of the strategy, by searching for our terms using the default keyword search in the Ovid interface. This maps search terms to a range of fields including titles, subject headings, heading words and abstracts. Finding the appropriate balance between sensitivity (recall) and specificity (precision) was achieved through the iterative process of refining and testing the search terms and strategy across the seven ‘core databases’.
Thesauri for the databases included in the review were checked to identify any further potential search terms and to ensure that all relevant subject headings were extracted to be included in the list of search terms. Encapsulating all relevant subject headings as search terms negated the need to perform exploded subject heading searches alongside mapping keywords to titles, subject headings and abstracts. The standard keyword search in the Ovid interface was used when searching, mapping the search terms to a default range of fields which included titles, subject headings, heading words and abstracts. This approach meant that it was unnecessary to develop separate search strategies for each of the ‘core databases’ built around their unique subject heading structure. A single, all-encompassing search strategy could be developed and applied across all of the databases in the Ovid interface. Pragmatically, this allowed for a single set of results to be generated from which duplicates could be removed by the Ovid ‘remove duplicates’ tool. Logistically, having a single final search strategy that could be used across the ‘core databases’ was a real bonus for this review, reducing the complexity of having multiple strategies that would need to be run individually.
Table 3 details the search strategy used on the various databases included in the review and the number of papers (n) retrieved on each database.
Table 3. Search strategy and results
|amed||Full search strategy as detailed in Table 2.||1895|
|BNI|| || |
|cinahl|| || |
|embase|| || |
|HMIC|| || |
|medline|| || |
| PsycInfo|| || |
| Sportdiscus (all on Ovid interface)|| || |
|Index to theses†||Limitations in search interface. Used search strategy of IT concepts combined with ‘health’. Gave n = 20|| 20|
|ISI Proceedings||Full search strategy as detailed in Table 2|| 70|
|MIDIRS||Full search strategy as detailed in Table 2 retrieved nothing—so reduced search to cover just IT concepts combined with attitude*|| 12|
|Zetoc||Limitations in search interface—can only use one term for each of our concepts—ran several of these through to ensure coverage of appropriate material—|| 17|
|Effectively search strategy was ... (information tech* or communication tech* or ICT or comput* or e health or ehealth or technology in health care or informatics or e learn* or elearn*) AND (attitud*)|| |
A data extraction sheet (Appendix 1) was developed to standardize the collection of information and enable rapid comparison between them.
The bibliographic details in the computerized database search including abstracts, where available, were either printed out or examined on the computer screen in a sifting exercise. For the sifting exercise, inclusion criteria focusing on research material examining health care personnel attitudes were agreed by the research team through ongoing discussions based on their experience and initial reviews of the literature. At this point, it appeared that there was thematic analysis being performed to further narrow down the material. A comparison of the selections made by the reviewers (1 : 20) was therefore undertaken independently to reduce selection bias.
Full papers were acquired if they fulfilled the criteria; however, even after this stage some items were excluded using the criteria shown in Table 4, as they were ambiguous, either containing limited relevance to the review or insufficient detail to meet all the criteria.
Table 4. Inclusion and exclusion criteria
|Included measures of attitude related to IT|
|Not in English|
|Not relevant to search aims|
|Not relevant to ‘health professions’|
|Not research (any form of original data collection)|
|Does not consider a concept related to attitudes|
|Case studies involving only one person/case|
|Single page or less|
|No author identified|
|Letters/personal viewpoints/reflective articles/opinion pieces/commentaries on papers unless clearly relevant and ‘evidence-based’ or collected as part of a properly designed study|
|Not pre- and post-registration/graduate education/continuing professional development (CPD)|
Even when explicit inclusion criteria have been specified, decisions concerning the inclusion of individual studies remain relatively subjective, as has been described in other studies.23 A number of papers that provided general analyses, literature reviews and other works were used to inform the review but were not included for detailed data extraction.
Each paper was read by one of the research team and the data extraction sheet completed, highlighting the setting of the study, the methodology used and the findings. The team then met and each contributed the ideas for themes which they had identified from the papers they had reviewed. An iterative process was undertaken to refine the themes and explore the overlaps and gaps between them. Each team member then identified which of the papers they had reviewed applied to one or more of the themes. A sample of the papers was then reviewed by a second member of the team to ensure consistency of allocation to themes.
A wide range of papers were reviewed, covering a variety of research methods and examining different aspects of the domain. The majority of the papers reflected quantitative and qualitative studies, mostly surveys, although there were several interview studies and a few observational studies with sample sizes reflecting the variations in methods.
The papers had collected data in 16 countries with 52% of those from the USA. Several studies were conducted before and after computerization of systems (e.g. Derup24), although the majority of studies were developed by the individual researchers, often with little evidence of piloting or checking the validity and reliability of the instrument. Studies were often conducted in conjunction with the introduction of new hardware or software systems into the areas in which respondents worked (e.g. Junger25).
The primary focus of the studies was the attitudes of staff to IT in general, but often they had a more specific or different focus and the data about staff attitudes were supplemental or tangential to the main study.
Information systems can be seen as an organized and coordinated approach by individuals to process and share information with each other. Therefore, an information system contains interfacing components, some of which are people, objects and procedures. The technical system is viewed as embedded in a social system—the human activity system includes the use of computer-based systems.26
The themes that emerged as affecting attitudes towards IT, ranged from the social system aspects of management—power, education and training, to the user-specific attributes of age, gender, professional groups and previous experience and also the design of the system. These themes are summarized in Table 5.
Table 5. Themes from the literature
|Gohlinghorst et al.27||Derup24|
|Chiasson & Lovato28||Lee35|
|Junger25||Fischer et al.36|
|Weir et al.29||Polhamus et al.61|
|Adams et al.30||Chan et al.55|
|Trivedi et al.31||Embi et al.62|
|Beuscart-Zephir32||Westerlund et al.63|
|Collins et al.33||Haluck et al.64|
|Pagliari et al.20||Sery-Ble et al.65|
|Design of software/hardware||Power and decision- making ability|
|Carroll & Christakis34||Stricklin et al.66|
|Lee35||Doolan et al.67|
|Fischer et al.36||Pare68|
|McAlearney et al.37||Chismar & Wiley-Patton69|
|Ammenwerth et al.38||Beuscart-Zephir32|
|Chiasson & Lovato28||Loomis et al.42|
|Johnston et al.71|
|Lai et al.72|
|Previous experience||Moffat et al.73|
|Liu et al.39||Araujo et al.74|
|Dixon & Stewart40||Temple et al.75|
|Moody et al.41|| |
|Ammenwerth et al.38||Telemedicine/imaging/ pathology systems|
|Loomis et al.42||Yousem & Beauchamp76|
|Balter43||Mast et al.77|
|Chan et al.55|
|Management/workload/efficiency||Lehoux et al.78|
|Rosenbloom et al.44||Demartines et al.79|
|Urkin et al.45||Bartoloni80|
|Rodriguez et al.46||Gattas et al.81|
|Apkon & Singhaviranon47||Crowley et al.82|
|Moody et al.41||Rydmark et al.83|
|Gadd & Penrod48|| |
|Kleiner et al.49||Professional issues|
|Leung et al.50||Krall & Sittig84|
|May et al.85|
|Professional groups||Seckman et al.86|
|Lee51||Ammenwerth et al.38|
|Porteus et al.52||Anderson87|
|Ammenwerth et al.38||Darr et al.56|
|Hersh et al.53||Campbell et al.88|
|Ash et al.54||Larcher et al.89|
|Chan et al.55||Travers & Downs90|
|Darr et al.56||Lee et al.91|
|Jacko et al.57||Liederman & Morefield92|
|Schubart & Einbinder93|
|Confidence in system privacy||Walter et al.94|
|Kouri et al.58||Age and gender|
|Darr et al.56|
|Decision support systems||Loomis et al.42|
|Bauer et al.59||Chan et al.55|
|Rosseau et al.60||Lai et al.72|
|Araujo et al.74|
Two types of barriers to the implementation of IT systems were identified by Leung et al.50 These were cost related (both financial and time/effort) and the knowledge/attitude of end-users. Incentives for the introduction of IT-based systems were either economical or driven through new legislation and regulatory requirements. One finding was that computer systems enable greater accountability and evaluation, giving managers more control over health care services.
Both Rodriguez et al.46 and Urkin et al.45 found that physicians highlighted the potential of computerized charts with office work, preventing the loss of information and facilitating communication. However, they had concerns about how patients’ records were displayed and anticipated an increase in workload and the time taken for data entry. Kleiner et al.49 also detected concern about workload issues in their study looking at e-mail consultations. In addition to being concerned with confidentiality issues, 80% felt that using e-mail would increase their workload. This was also supported by Gadd and Penrod,48 who found the time taken for documentation was increased by a computerized system. Contrary to this, Moody et al.41 found that more than one-third of nurses (36%) perceived that EHRs had resulted in a decreased workload.
Other impacts in addition to workload were identified by Moody et al.,41 who found that respondents believed that electronic charting would lead to improved safety and patient care, whilst another study found improved efficiency.44 Electronic documents contained 50% more descriptors than handwritten documents.47 Colleagues derived benefits from increased legibility, more detailed content and the ability to generate administrative and patient-care reports.
The majority of nurses (64%) preferred bedside documentation, but also found that environmental and system barriers often prevented EHR charting at the bedside. EHRs are seen to improve the quality of documentation.41
Power and decision-making ability
Stricklin et al.66 found that nurses’ work and the barriers they perceived to making changes each appear to explain more variance in nurses’ attitudes toward computers than security issues; and that it is vital for the industry and its providers to be prepared for the myriad ways point-of-care (POC) technology will affect the practice of nurses, the end-users. The ultimate satisfaction and use of the technology of home-care nurses will greatly affect not only the business success of the organization but also the structure, process and outcomes of care to patients and their families.
The various methods that staff use to challenge the implementation of IT systems was studied by Timmons,70 who found that resistance took a wide variety of forms, including attempts to minimize use of the systems and extensive criticism of the systems, although outright refusal to use them was rare. Resistance was as much about the ideas and ways of working that the systems embodied as it was about the technology being used. The patterns of resistance can best be summed up by the phrase ‘resistive compliance’. Another aspect of non-compliance was discussed by Beuscart-Zephir,32 who found that, whilst novice anaesthetists followed the system, experienced anaesthetists jumped around the questions, underlining and setting alerts, etc., surrendering the system useless or depersonalized.
Chismar and Wiley-Patton69 used the Extended Technology Acceptance Model (TAM2) to examine paediatricians and showed perceived usefulness was a strong determinant of intention to use, whereas perceived ease of use and subjective norm did not significantly affect intention to use IT. Further analysis identified job relevance and results demonstrability as factors determining perceived usefulness. Physicians are willing to adopt beneficial applications in IT, even if they are not easy to use; the tool used (TAM2) did not take into account the higher level of competence, intellectual and cognitive capacity that physicians generally have, also the ability to adapt to new technologies.
The success of implementation of clinical information systems was studied by Doolan et al.67 The presence of high-level leadership was considered the single most important factor contributing to successful implementation. Each site had the aim of improving quality of care through implementation of electronic records, with emphasis placed on the clinical processes being supported by the system and not driven by them. The study also identified that it is important to build on the momentum of earlier successes to achieve widespread implementation and use of a new system. This is also supported by Johnston et al.,71 in that negative attitudes towards computerization were only displayed by physicians that do not have an existing IT system. The majority of negative attitudes are related to the cost of computerization.
Organizationally successful implementation was about identifying champions and getting the right people on board (see Ash et al.54). The selection of the implementation strategy depended on the experience, skills, beliefs and motivation of the key actors involved. Pare68 concluded that health care IT projects are always characterized by some vagueness and projects can never be perfectly controlled or predicted.
Education and training
The potential of IT in facilitating training has been explored in a number of studies. Haluck et al.64 suggest that programme directors believe this type of technology would be beneficial but lack adequate information regarding virtual environment simulators. This was supported by Polhamus et al.,61 who examined the potential for the use of technology in distance education and concluded that, to make effective use of web-based resources, both technology skills and confidence in using those skills are essential. It should not be assumed that those enrolling in distance learning have the necessary technical skills to complete training successfully.
In the study by Sery-Ble et al.,65 nurses believed that the computerized presentation was an effective way to learn new material, even though only 57.1% considered themselves proficient with computers. They suggested future studies should also examine the types of medically related skills that can be effectively taught by computerized training programmes. This is supported by Zdanuk et al.95 who showed that computer-based learning is beneficial for the confidence of clinicians.
Searching skills and access to adequate technology was highlighted by Moffat et al.73 as areas stopping clinicians using online information effectively. Despite general positivity, Hancock et al.96 also found a need for more computer literacy, information literacy and research education.
Comparisons between professional groups were often difficult. Darr et al.56 found nurses were the most enthusiastic about IT implementation, while junior doctors were the least. The junior doctors resented the system as they had to enter most of the data, whereas nurses stressed the positive impact of the electronic medical record (EMR) on improving patient care. Porteous et al.52 found that, whilst general practitioners (GPs) had concerns about technical issues, pharmacists were more worried about workload issues.
In another study, hospital physicians felt the system to be technically cumbersome and time-consuming, and in the teaching hospitals resented using the ‘physician order entry’ (POE) system. However, in a community hospital setting, staff had been involved in developing clinical pathways, resulting in a system which had a positive organizational impact as it was multi-disciplinary.
In a study of a range of health professions, Jacko et al.57 showed significant results as a result of the occupation of the participant. The nature of the occupations influenced the need for a more comprehensive list of possible search subjects and topics and different search tools. Nurses need access to more sites than pharmacists and specialists, although nursing students were the most likely to be non-users of the Internet.
The fact that the requirements of the different professional groups vary was also documented by Ammenwerth et al.38 Medics welcomed speedier access to information, recording data at the patients bedside and the ability to communicate anywhere—leading to health care professionals working together to provide better and improved patient care. However, Lee51 found that, whilst potential users of the ambulatory care EMR had generally positive or neutral attitudes toward the system, physician groups had less positive feelings than the other professional groups.
Social network analysis can be used to analyse relationships between health care providers and to identify influential individuals who are critical to the successful implementation of IT systems.87 Identifying barriers to successful implementation was the focus of some studies. Campbell et al.88 concluded that implementation strategies should be tailored to the environmental conditions of the practice sites. In further detail, Darr et al.56 identified six main domains of concern expressed by physicians: managerial implications of EMR; limits on professional autonomy; its impact on communication with colleagues; facilitation of research; legal defence; and influence on the professional hierarchy within the hospital. Campbell et al.88 also found six similar and related themes examining care providers’ receptivity to technological change: turf; efficacy; practice context; apprehension; time to learn; and ownership. Care providers and administrators consider a range of factors, including economic ramifications, efficacy, social pressure and apprehension, when deciding whether and how fast to adopt telemedicine. Travers and Downs90 also found five main themes. These were: usefulness; organizational culture; relationships and partnerships; and the challenge of implementing a system developed for an academic centre environment into private practice settings.
Once implemented, user commitment and attitude change; Krall and Sittig84 identified five main topics in their focus groups—efficiency, usefulness, information content, user interface and workflow being key issues. The focus groups held generated some innovative suggestions and surprised the researchers with the level of emotional engagement.
Lee et al.91 found that intensive care unit (ICU) nurses were ambivalent towards a new system as some shortcomings of the manual processes were not addressed by implementing the system. However, major themes were: system saves resources; individualizes care plans; waiting for printouts from system; routine based; and there was no consensus on nursing diagnostics, concluding that there was a shift for nurses from being patient centric to information systems centric.
Although physician productivity increased significantly after introduction of the system, Liederman and Morefield92 identified speed of the computers and extra workload as reasons for a negative attitude towards the system. Seckman et al.86 found a strong correlation was detected between adoption of technology and perceived usefulness, as well as significant differences between specialities being identified about ‘perceived usefulness’ and ‘impact’, but no differences were found in relation to perceived ‘ease of use’.
Schubart & Einbinder93 identified the importance of organizational culture and the need for data was illuminated by the executive management interviews. Their study also found that compatibility with an individual's work style and skills was associated strongly with satisfaction and continued use of the system. Improvement in care and communication was also documented by Walter et al.94
Findings by Dixon and Stewart40 showed that, amongst primary care physicians, usage of IT showed significant differences among groups for intent, interest, perceived usefulness, perceived ease of use, finesse and knowledge. This was also supported by Moody et al.41 who found that, amongst nurses with expertise in computer use, 80% had a more favourable attitude toward EHRs than those with less expertise. Also, positive correlation between acceptance of the system and years of experience of computer systems was demonstrated by Ammenwerth et al.38 Users with less positive attitudes had previously worked in areas with less thorough documentation procedures prior to the implementation of the system, therefore, post-implementation, the nurses had experienced an increased workload in order to supply complete documentation. Possible links between the number of motivated key users and department-wide acceptance of the system was also explored. Acceptance increases over time as users become familiar with the system. In a follow-up study,38 years of computer experience is positively correlated to initial acceptance of computer use and it was concluded that the fit between the task and the selected technology is important for user acceptance. Poor fitness for use can have a negative effect on quality of patient care provision.
A study by Balter43 found attitudes to computers and e-mail became more positive over the years, especially amongst the employees. The majority of managers were positive at the start of IT implementation and remained positive.
Location has been demonstrated to impact on attitude to IT; Loomis et al.42 found EMR non-users were more likely to practice in a suburban or rural location. Non-users were significantly less likely than users to believe that (i) physicians should computerize their medical records; (ii) current EMRs are useful; (iii) EMRs will reduce their risk of making medical errors; and (iv) EMRs will improve health care quality in their office.
Liu et al.39 found that nurses’ computer skills were significantly and positively correlated with both computer knowledge and computer attitudes; however, no significant correlation was found between computer knowledge and computer attitudes.
Age and gender
Although it was anticipated from other literature that age and gender were significant factors in attitudes to IT, with female and older users being less positive, papers in this review generally did not find this to be the case (e.g. Chan et al.55).
Lai et al.72 were unable to demonstrate any influence of age and gender on the translation of intention to actual implementation of a system. Loomis et al.42 found that there were no statistically significant differences in age or gender between users and non-users of EMR systems. Although, in another study, Moffat et al.73 found 41% of female GPs were non-users compared with 28% of male GPs, Araujo et al.74 found that computer use, age group and gender were not significant in explaining the attitude. They had, however, found evidence of a strong intention towards utilization and these intentions are conditioned by attitude.
Confidence in system privacy
Poor content design, system function and system integration, led to the nurses seeing the implementation as a policy requirement, rather than enhancing care and they were further worried about privacy and legal issues. Lee35 found that nurses’ interest in IT was limited because of concerns about inconvenient access to computers; reduced work efficiency; inability to individualize patient care and nursing speciality deficiencies.
Gadd & Penrod48 found that partly as a result of worries that they would have an even more detrimental effect on personal and professional privacy than they had previously judged. Physicians felt disenchanted with the Electronic Medical Record (EMR). However, Loomis et al.42 established that users consider EMRs as more secure and more confidential than paper based records. Supporting this, all participants in a study by Kouri et al.58 considered adequate privacy protection as a prerequisite for development of a system.
Design of software/hardware
Ammenwerth et al.38 described how one project was required to support ‘hot line’ initially, because staff were handling new and complex hardware and software. Therefore they were unable to demonstrate reduced costs related to the implementation of this new technology.
Carroll et al.34 were among many studies which either directly looked at the design of software and hardware for use in health care settings or produced comments from participants which related to these areas. Lee,35 amongst others, found hardware availability, content design and user training/education programmes are critical issues that affect nurses’ use of computers in their daily practice. Effect of hardware availability was also underlined by McAlearney et al.37
The importance of flexibility in software design and project implementation was also emphasized by Beuscart-Zephir.32 Users demonstrated that they needed to customize the options that were available within the software, based on their professional judgement. Software compatibility with factors related to implementation, user background and characteristics of the setting are important in influencing use. Software should be evaluated for its capability to induce change in work practices in order to achieve both short- and long-term relative advantage.28
Clinical records (EMR)
Whilst Weir et al.29 found overall user satisfaction was relatively low on all measures and the level of adoption was quite low, none of the users interviewed would return to paper-based systems. The need for good communication in software development was emphasized by Trivedi et al.31
A well-designed EMR system can provide positive benefits. Adams et al.30 found paper-based visit records had more content, whereas computer-based visit records covered more areas and provided improved risk assessment. Although, despite them liking the computer charting better than the paper records, Gohlinghorst et al.27 found nurses wanted a record of the patient condition over time. Notwithstanding this, Collins et al.33 found more comprehensive information was stored per episode than with paper-based systems. They also reported that patients found it easy to use, but health professionals had a more complex interface and found the system more difficult to use.
Despite positive attitudes to the system being found by Junger,25 this study also raised some questions about the hardware and software and recommended that users should be trained to cope with problems and computer ‘bugs’ inherently present in today's highly complex software environments.
Chiasson and Lovato,28 in their study of users views about new software for health care practice, obtained user comments which reflected all five of Rogers’97 perceived characteristics of an innovation, i.e. relative advantage, compatibility, trialability, complexity and observability. The software slowed down project momentum and created a reflective planning group when an action-orientated group was needed.
High levels of computer familiarity, welcoming attitudes and positive ratings of usability, format and utility were revealed in a study by Pagliari et al.20 Content analysis of electronic feedback revealed mainly technical queries and general expressions of satisfaction.
The implementation of a Picture Archiving and Communication System (PACS) in imaging departments had the most specific quantitative documentation,55 identifying 91% in productivity gains; 70% of radiologists reported improved diagnostic capabilities. Allied health professionals (AHPs) productivity gained by 50% on average. Despite this, Yousem and Beauchamp76 highlighted the importance of consulting ‘customers’, i.e. surgeons rather than radiologists only. Additionally, Mast et al.77 found utilization of the system appeared to be highly dependent on the individual rather than speciality. Challenges included mapping procedural skills from one medium to another and general confidence in the quality of the images and tools on the system. This is supported by the work of Crowley et al.,82 who studied the use of pathology images in the electronic record and found that image quality and issues of access and interpretation needed further evaluation.
Satisfactory quality of images and audio, even improvements to that which existing facilities offered was documented by Rydmark et al.83 and Gattas et al.81 Bartoloni80 evaluated the ease of setting up systems in remote locations, reflecting an important factor in system design. Another study found that analysis of existing structures will improve uptake by system users. Lehoux et al.78 found that technology does not always reflect existing structures and use of technology reflects fitness for purpose and concluded that software designers often impose structures on users.
Decision support systems
Mixed responses were found with systems designed for decision support. Bauer et al.59 found that staff were generally satisfied with the system and had discovered that it was of some benefit to their practice, whilst Rosseau et al.60 found negative comments about the decision support system significantly outweighed the positive or neutral comments. Respondents did not believe that the system fitted well within the general practice context, with concerns emerging in three main themes—timing of the guideline trigger, ease of use of the system and helpfulness of the content.
This review aimed to identify relevant literature, review and summarize the methodologies and findings and make recommendations for future practice and research in the area.
The process of identification, selection and review of the articles included in the study was complex and time-consuming. The initial search was carried out in late 2005 and therefore more recent articles were not included. The keywords and databases used gave a wide coverage and, while not guaranteed to be comprehensive, do give a reflection of the information which has been published in this area. Data extraction and thematic analysis is a subjective process based on the knowledge and interests of the research team, but with a series of checks and protocols in place to make this as systematic as possible.
A wide range of material was reviewed from around the world which revealed some high degrees of consistency amongst the studies, but also findings that were unique to a particular country or health care system.24,50 This added some interesting ideas and views to the review. For example, the Dørup25 study of medical students in higher education institutes (HEIs) coping with IT literacy and ownership of computers found that, in the year 2000, 79.7% males and 67.5% females in Denmark, compared with 44% males and 37% females in Edinburgh, had their own computers at home. This identifies both gender differences and resources issues that are likely to impinge on attitudes towards IT. Similarities between countries were noted from the study of Leung et al.50 when researching barriers and incentives to the use of clinical computing in Hong Kong identified that time costs, lack of technical support and capital investment were barriers. Improved office efficiency and patient care were also incentives amongst the practicing physicians. Major country differences could be seen in the ranking of software copyright privacy regulations and patient, public and regulatory demands for computer systems were ranked as least barriers and as catalysts for action. Physicians had differing priorities, from concerns with cost to patient-related factors to personal/physician issues. Leung et al.50 pointed out that some countries, for example, the UK, were governed by legislation, whilst countries such as Canada required all physicians to electronically register fees for their services. It was suggested that Hong Kong should also look to government regulations and financial assistance to improve the situation in practice areas.
Much of the literature reviewed used questionnaires such as the data collection method, although interviews, focus groups and observation were also represented. In very few cases49,92 were attempts made to measure the effects on patients, although this may sometimes be inferred by the views of the health care professionals involved. This is an area that requires further research.
Overwhelmingly, the issues that affected health care professionals’ use of computers in their daily practice were hardware availability, content design and user training/education programmes.28,35,39,41,55,71,94 The specific computer hardware and software programs varied in the papers, but, for the majority, there were consistent issues about the fitness for purpose and usability. This suggests that there are still a large number of places and health workers that are not ready for IT.
Many of the studies included in the review related to the introduction of electronic patient records,23,29,30,41 reflecting their increasing development and use during the time period, although studies examining telemedicine, pathology and radiography systems were also included.78,80,81 Many users were willing to overcome some of the difficulties of implementation if they saw long-term benefits. This focus amongst the studies included in this review may reflect a particular point in systems life cycles and different views about the use of IT may be found if more studies were conducted at different stages in the process and if longitudinal studies were carried out over a number of years.
A variety of studies considered the education and training needed for successful use of health care IT systems64,65 whereas others looked at the use of IT in more general education and training.61,62 Comments included the needs expressed by users both for more education and training before the implementation of IT-based work systems and for more on the use of e-learning approaches to find out about professional health care practice.
Experienced IT users tended to have more positive attitudes towards the introduction and use of IT in their work setting.38 A more significant factor in many of the studies was the relationship between the IT system and the user's professional roles and power relationships, with some individuals seeing challenges to their role and traditional practice, as seen in the study of Timmons70 when he refers to ‘resistive compliance’. Explanation of the value of IT systems is therefore a necessary part of staff preparation when planning implementation or change. In a survey by Dillon et al.98 it showed that experienced and mature nurses were becoming more compliant and eager to utilize IT resources, thus reinforcing the research evidence. Attitudes of some professionals have changed greatly, but other professional groups have still to accept the changes.
Some issues were frequently cited concerning the introduction and use of new IT systems and their effects on workload and efficiency and perceived issues around privacy and confidentiality.58 Particular issues arose within the first 6 to 12 months of implementation when the change is hardest and efficiency gains had yet to be obtained. In 1995, Newton99 had highlighted that nurses attitudes to IT systems, although negative at 6 months, were less negative after 1 year of use in the practice setting, with a subsequent improvement in quality-of-care planning. There still appears to be some continuation of efficiency and data security concerns as the literature shows (see Moody et al.41). A more recent study in 2007 by Lee100 in Taiwan states that, although the same concerns were present, the nurses’ attitudes were positive. The answer appears to emphasis adequate education before implementation.
Some of the major issues identified related to the drivers for implementation, which were often ‘top down’ and the implementation at lower levels in the organization. Generally, those systems with a clear design and which build-in education and training were conceived to be the most successful. The previous IT experience of the users was also important in influencing implementation.
The wide variety of techniques used for data collection and analysis made comparisons between the different studies and settings difficult and, in many cases, it was hard to assess the quality and rigour of the studies. The use of single instruments, which had not been piloted or tested for validity and reliability, means that the reliance which can be placed on them in the development of policy and practice is limited. The implementation of IT systems into health care delivery organizations is increasing and, if some of the individual and cultural issues are to be addressed, there is therefore a need to commission high-quality research which uses validated instruments to examine the attitudes of staff. In particular, independent studies which identify effective strategies to change or manage staff attitudes to IT are needed to underpin future development and implementation.
There have been significant changes and investment in the UK's NHS in relation to IT over the last 5 years and many of the issues found in this review may have already been addressed. However, it is likely that, as issues around IT implementation are becoming concerns for a range of managers at practice level, the factors identified could be argued to be more relevant than ever. Those responsible for implementation at national and local levels will need to take the findings into account and invest time and effort into strategies that are likely to produce more positive attitudes amongst the workforce.
The issues identified in this review are significant for practitioners, researchers and students in the library and health professions, as the increasing use of IT for the identification and retrieval of the evidence to support practice requires adaptation of the processes and education of end-users, taking into account the attitudes of those staff. Educational programmes need to be adapted to the individual traits of the users, but also to their role within the organization and professional status. IT system development needs to take into account the flexibility required for use in a variety of clinical settings as well as in the library or home environment. Extensive testing and user profiling at all stages of development and implementation is needed to ensure that health care staff are able to quickly see the real benefits of IT-based systems supporting evidence-based care.
This study provides a comprehensive review of the literature related to health care staff attitudes towards IT, providing an overview of current developments and identifying key factors which influence these attitudes.
A range of key issues, such as the need for flexibility and usability, appropriate education and training and the need for the software to be ‘fit for purpose’, showed that organizations need to plan carefully when proposing the introduction of IT-based systems into work practices. Also, those organizations could benefit from involving users with higher levels of previous experience of IT, as they tended to have more positive attitudes. Education in the use of IT for health care professionals at undergraduate and postgraduate or continuing education levels is therefore an important aspect. These findings need to be taken into account in future programme development and implementation.
The studies reviewed did suggest that attitudes of health care professionals can be a significant factor in the acceptance and efficiency of use of IT in practice. However, there do not appear to be any consistent indicators of their likely attitude other than experience and confidence in using IT.
The quality of the studies and the wide variety of methods employed mean that it is difficult to place great reliance on their findings or identify commonalities or variations, whether based on geography or context. Although the themes identified do seem to recur in many studies, and therefore may well be significant for practice, further qualitative and quantitative research is needed into the approaches which have most effect on the attitudes of health care staff towards IT. These may well include quantitative approaches such as attitude scales and estimations of time taken to achieve particular tasks, but also qualitative studies which try to explore in greater depth the varied factors which form and influence attitudes. Consideration can then be given to the strategies which are likely to be most effective. This may include the use of the IT systems found in practice settings in library and educational environments. As ever greater financial and other resources are allocated to IT systems in health care, the factors which influence staff attitudes towards them become increasingly significant if the investment is to be worthwhile. If information systems underpinned by new technologies are to play a significant role in the expansion of evidence-based practice, then the human factors of those who will use them needs to receive as much consideration as the technologies themselves.