Professional practice and innovation: users' attitudes to an electronic medical record system and its correlates: a multivariate analysis.
Enhancing service efficiency and quality has always been one of the most important factors in heightening competitiveness in the healthcare service industry (Gillies et al. 2006; Gold, Glynn & Mueser 2006). With the increase in the demand for high quality medical services, the need for an innovative Hospital Information System (HIS) has become essential (Yoo et al. 2008). HIS and electronic medical records (EMR) are considered prerequisites for the efficient delivery of high quality care and instrumental to the decrease in medical errors in healthcare delivery (Wang et al. 2003; Lium, Tjora & Faxvaag 2008). It is widely acknowledged that the EMR has the potential to become the core electronic information and communication system in the healthcare sector (Ball 2003; Haux 2006; Chang & Chang 2008). The EMR eliminates the need for handwritten records, hence reducing time spent on record-keeping by doctors and other healthcare staff, which in turn results in reduced patient waiting time (Cimino et al. 2002). In addition, tracking time on historically-related records is also reduced simply by accessing the computerised medical records through the integrated computerised central system (Sharda et al. 2006). EMRs may also increase both patient satisfaction and financial savings (Sittig, Kuperman & Fiskio 1999).
With the assumption that that EMR system can improve both the quality and effectiveness of the healthcare delivery, many healthcare provider organisations in developed countries have invested in the development and deployment of such systems (Lium et al. 2006). However, in many developing countries the EMR system is not widely disseminated or implemented. Published literature shows low adoption and high failure rate of successful EMR implementations (Southon, Sauer & Dampney 1999; Benson 2002; Ball 2003; Littlejohns, Wyaat & Garvican 2003; Ash et al. 2004). Van der Meijden et al. (2003) identified user resistance as one of the primary factors for unsuccessful EMR implementation. Despite their potential advantages, implementation of EMR systems may be resisted if users are not satisfied with the system (Van der Meijden et al. 2003). Although there is evidence of immediate benefit (Berg et al. 1998) and the improvements in time efficiency (Poissant et al. 2005), other factors, such as deficiencies in users' computer skills (Dansky et al. 1999; Van der Meijden et al 2001; Ammenwerth et al. 2003) and the implementation process (Southon et al. 1999; Aarts, Doorewaard & Berg 2004; Berg 1999), also had a negative impact on successful EMR implementation.
Previous published work has identified many factors that may influence EMR user satisfaction and ultimately help its successful implementation. These are system error, response time, logical and efficient flow of tasks, ability to complete desired tasks, ease of correcting mistakes and data entry, effects on individuals' time, prior computer experience and training on the system (Dansky et al. 1999; Aaronson et al. 2001; Murrf & Kannry 2001; Loomis et al. 2002; Likourezos et al. 2004). Van Der Meijden et al. (2003) also suggested that perceived impact on the quality of patient care may be an important factor in users' satisfaction with an EMR, especially when the system is perceived as negative.
Kuwait is a small oil-rich Arab-Muslim country of 3.2 million people, only 33% of whom are Kuwaiti nationals. Non-Kuwaitis are from over 100 countries; 28.7% are Arabs and the rest (33.3%) are from various other Asian countries (PACI 2007). The healthcare delivery system in Kuwait has developed rapidly. The Ministry of Health of the government of Kuwait provides about 90% of the healthcare service through the usual three-tier healthcare delivery system, comprising primary, secondary and tertiary levels. As stated in 1999-2003, one of the objectives of the Ministry of Health in Kuwait is to establish an EMR system in all primary healthcare centres, under the Ministry's computerisation program. The Ministry aims to improve the quality of healthcare by establishing a modern information technology system in all related processes and extending the system to various levels of healthcare (Ministry of Health 2004). It believes this will improve health professionals' performance, and hence the quality of patient care; it is also consistent with the Kuwait government's overall plan to have an 'electronic government'. At present, EMR has been implemented in all primary healthcare centres.
Evaluation of user interaction and satisfaction with the EMR will provide important information that will allow the user to have a greater voice in design and delivery of healthcare services at the level of primary healthcare. Findings from the primary healthcare centres' experience will help policy makers to plan future implementation of HIS at the secondary and tertiary care levels. This study was aimed at identifying receptionist's reaction to the newly implemented EMR systems at primary healthcare centres in Kuwait and its correlates.
At the time this research took place there were 877 medical receptionists in total who were using computerised systems in 78 health centres in the five health regions of Kuwait. During the first stage, 44 of these 78 centres were selected, covering more than 50% of patient population. A sample of 400 receptionists was randomly selected for the interview; they returned 332 completed questionnaires, a response rate of 80.5 %. The data collection period spanned November to December 2005.
A structured questionnaire was prepared to measure the user reaction regarding the computerised system at the health centres. A user interaction satisfaction questionnaire (QUIS) developed by the researchers from the Human Computer Interaction Laboratory at the University of Maryland was adopted with minor modification (e.g. to some of the wording) for the current study. It was a standardised, general user evaluation instrument for interactive computer systems. Psychological test construction method was used to develop this questionnaire to ensure reliability and construct and empirical validity. The short form of the QUIS was divided into five sections of four to six questions each. The sections were designed to assess: (a) overall user reactions; (b) screen design and layout; (c) terms and system information; (d) learning; and (e) system capabilities. In order to ensure its validity, the questionnaire was translated into Arabic and back-translated by two Faculty members who were fluent in both English and Arabic. The Arabic version of the questionnaire was used, since most of the receptionists were only Arabic speaking. Necessary permission was obtained from the Ministry of Health before conducting the survey. Respondents were informed that participation in this survey was voluntary.
Some questions related to background characteristics (e.g. age, gender, education, nationality and duty shift) and computer literacy-related questions (previous computer experience, possession of computer at home and office, daily computer use frequency, typing ability, ease of data input and computer error frequency) were also included.
Six questions (yes=1 or no=0) about overall user reaction regarding EMR were asked. Respondents were asked whether they found their system to be: (a) rigid; (b) adequate; (c) difficult; (d) dull; (e) frustrating; or (f) very bad. Overall user reactions scores were computed by adding responses to all six questions. The minimum value for computed overall user reactions score is 0 (no negative reaction) and the maximum 6 (maximum negative answer). For bivariate analysis, overall user reaction score were recategorized as 0, 1, 2 and 3 (3 or more negative answers). For Poisson regression, this is used as a count variable.
Data were analysed using SPSS. Descriptive statistics (frequency distribution and bivariate table) were used to present the data. A chi-square test was used to determine the significance of the relationship between two categorical variables. In addition, Poisson regression was used to identify independent correlates which influence user reaction to the EMR system.
Overall user reactions regarding the various aspects of the EMR system are presented in Table 1. Distribution of the background characteristics by overall user reactions are presented in Table 2.
Among the respondents, 23% were aged 25 years or less, 43% between 26 to 35 years, and 34% were 36 years or over. The relationship of age with overall user reaction was found to be significant: 62% of respondents were female and 38% male. Regarding education, 28% completed intermediate schooling (5th to 8th grade), followed by 31% secondary education (to 12th grade), and 41% college or higher degree. A majority of the respondents (98%) were Kuwaiti. Approximately 68% of the respondents were from the morning shift of work duty while the remaining 32% were from the evening shift.
Table 3 presents the data on computer-related experience of the respondents by overall user reactions. Of all respondents, 71% had a computer at home and 68% had previous computer experience. A majority of the respondents (85%) had a computer at work. Regarding typing ability, 49% reported as good, 39% moderate and 13% bad and the relationship between the typing experience and overall user reaction regarding EMR was significant. Of responses related to the ease of data entry, 90% reported it as easy and remaining 10% found it difficult. Forty-eight percent of respondents mentioned rare computer error, followed by, 48% sometimes and 4% always.
Logistic regression was employed to determine which characteristics independently correlated with overall user reaction. For the multivariate analysis, all variables from Tables 2 and 3 were included in the logistic regression model (Table 4). The dependent variable was counted as mentioned in the Method section. Two variables, ease of data input and computer error frequency, emerged as significant correlates with overall user reaction. Overall negative reaction was significantly lower for those who felt data entry was easy. Similarly, respondents who reported seldom/sometimes computer error showed significantly lower negative reaction regarding EMR.
This study investigated the overall user reaction and its correlates concerning the newly implemented EMR at primary healthcare centres in Kuwait. A majority of the respondents found the system flexible (83%), easy (89%) and satisfying (81%). However, more than one-third of the respondents reported that the system was inadequate (38%), and very bad (36%). Furthermore, a half of the respondents perceived the system to be dull. According to published literature, increased positive outlook on these aspects of EMR assists its increased use and successful implementation (Laerum, Karlsen & Faxvaag 2004; Lium et al. 2006; Lium et al. 2008). The medical receptionists' tasks are usually limited in scope, and have a narrower and more defined range of the type of information needed compared with those for nurses and physicians. Hence, their tasks could be incorporated in the EMR system in a simpler way, leading to a more favourable user reaction, and ultimately leading to its successful implementation.
In our study, bivariate analysis showed that age had a significant positive relationship with the overall user reaction. This might be because older people are reluctant to accept new technology, particularly computer systems, as has been observed in previous studies (Marasovic et al. 1997; Simpson & Kenrick 1997; Alquraini et al. 2007).
Respondents with moderate to good typing ability reported significantly fewer negative views on EMR. A similar significant trend was shown for ease of data entry and computer error frequency with respect to overall user reaction. A negative response to these variables suggests diminished user satisfaction and influence on EMR implementation. Published reports show a mixed picture regarding these relationships (Murff & Kannry 2001; Loomis et al. 2002; Likourezos et al. 2004; Laerum, Karlsen & Faxvaag 2004; Otieno et al. 2008; Lium et al. 2008). In the present study, for example, the significance of the impact of typing ability did not remain so in multivariate analysis. Although we found several significant associations in bivariate analysis, the association between two variables does not necessarily imply a significant causal relationship between them. Therefore, a multivariate approach was applied to determine which factors best explain and predict overall user reactions.
In this study, multivariate Poisson regression identified the ease of data input and computer error frequency as significant correlates of overall user reactions. As medical record receptionists enter basic Master Patients Index (MPI) data and other administrative data, the entry screen should be simple and easy to use. Laerum, Karlsen and Faxvaag (2004) argued that the EMR should be developed in such a way that medical receptionists can perform their tasks easily. This will improve receptionists' approach to EMR and will ultimately help in successful implementation of the system.
Computer error can occur for several reasons, namely hardware failure, network failure and any problem with EMR. Any breakdown of the EMR system has to await the arrival of the specialist technicians and during that time the user reverts to the paper-based system. This generates frustration and disappointment with EMR. Haux (2006) reported that the combined use of both EMR and paper-based systems will reduce the chance of successful implementation of EMR.
Proper training of EMR users will increase their computer skill and confidence with the use of the system. This will also help them to solve some simple computer and software-related problems. Ultimately, this is likely to enhance users' positive reactions towards EMRs. Lium et al. (2006) suggested that factors such as computer skills, training, and a well-functioning system are necessary but not adequate for success.
Limitations of the study
The present study included receptionists from government primary healthcare centres only, which limits the generalisability of findings to secondary and tertiary care facilities. In addition, the survey instrument could have been broadened to encompass some other specific areas of user satisfaction with the EMR system.
This study showed that medical receptionists' satisfaction with EMR depends on ease of data input and reduced computer error frequency. Therefore, it is recommended that concerned decision makers ensure that technical help is available promptly, when required. This study provides a set of baseline data and some pointers to conducting future studies with larger samples from different sectors of healthcare. The findings should also help policymakers to recognise the causes of dissatisfaction with the EMR system among medical record receptionists that adversely affect its successful implementation, regular use, and quality of care at the clinics.
Aaronson, J.W., Murphy-Cullen, C.L., Chop, W.M. and Frey, R.D. (2001). Electronic medical records: the family practice resident perspective. Family Medicine 33: 128-32.
Aarts, J., Doorewaard, H. and Berg, M. (2004). Understanding implementation: the case of a computerized physician order entry system in a large Dutch university medical center. Journal of the American Medical Informatics Association 11(3): 207-216.
Alquraini, H., Alhashem, A.M., Shah, M.A. and Chowdhury, R.I. (2007). Factors influencing nurses' attitudes towards the use of computerized health information systems in Kuwaiti hospitals. Journal of Advanced Nursing 57(4): 375-381.
Ammenwerth, E., Mansmann, U., Iller, C. and Eichstadter, R. (2003). Factors affecting and affected by user acceptance of computer-based nursing documentation: results of a two-year study. Journal of the American Medical Informatics Association 10(1): 69-84.
Ash, J.S., Gorman, P.N., Seshadri, V. and Hersh, W.R. (2004). Computerized physician order entry in U.S. hospitals: results of a 2002 survey. Journal of the American Medical Informatics Association 11(2): 95-103.
Ball, M.J. (2003). Hospital information systems: perspectives on problems and prospects, 1979 and 2002. International Journal of Medical Informatics 69(2-3): 83-91.
Benson, T. (2002). Why general practitioners use computers and hospital physicians do not--Part 1: incentives. British Medical Journal 325(7372): 1086-1094.
Berg, M., Langenberg, C., vd Berg, I. and Kwakkernaat, J. (1998). Considerations for sociotechnical design: experiences with an electronic patient record in a clinical context. International Journal of Medical Informatics 52(1-3): 243-251.
Berg, M. (1999). Patient care information systems and health care work: a sociotechnical approach. International Journal of Medical Informatics 55(2): 87-101.
Chang, H.H. and Chang, C.S. (2008). An assessment of technology-based service encounters and network security on the e-health care systems of medical centers in Taiwan. BMC Health Services Research 8: 1-13.
Cimino, J.J., Patel, V.L. and Kushniruk, A.W. (2002). The patient clinical information system (PatCIS): technical solutions for and experience with giving patients access to their electronic medical records. International Journal of Medical Informatics 68: 113-127.
Dansky, K.H., Gamm, L.D., Vasey, J.J. and Barsukiewicz, C.K. (1999). Electronic medical records: are physicians ready? Journal of Healthcare Management 44: 440-5.
Gillies, R.R., Chenok, K.E., Shortell, S.M., Pawlson, G. and Wimbush, J.J. (2006). The impact of health plan delivery system organization on clinical quality and patient satisfaction. Health Services Research 41: 1181-1199.
Gold, P.B., Glynn, S.M. and Mueser, K.T. (2006). Challenges to implementing and sustaining comprehensive mental health service programs. Evaluation and the Health Professions 29: 195-218.
Haux, R. (2006). Health information systems--past, present, future. International Journal of Medical Informatics 75: 268-281.
Laerum, H., Karlsen, T.H. and Faxvaag, A. (2004). Use of and attitudes to a hospital information system by medical secretaries, nurses and physicians deprived of the paper-based medical record: a case report. BMC Medical Informatics and Decision Making 4(18): 1-10.
Likourezos, A., Chalfin, D.B., Murphy, D.G., Sommer, B., Darcy, K. and Davidson, S.J. (2004). Physician and nurse satisfaction with an Electronic Medical Record system. The Journal of Emergency Medicine 27(4): 419-424.
Littlejohns, P., Wyatt, J.C. and Garvican, L. (2003). Evaluating computerised health information systems: hard lessons still to be learnt. British Medical Journal 326(7394): 860-63.
Lium, J.T., Lerum, H., Schulz, T. and Faxvaag, A. (2006). From the front line, report from a near paperless hospital: mixed reception among health care professionals. Journal of the American Medical Informatics Association 13(6): 668-675.
Lium, J.T., Tjora, A. and Faxvaag, A. (2008). No paper, but the same routines: a qualitative exploration of experiences in two Norwegian hospitals deprived of the paper based medical record. BMC Medical Informatics and Decision Making 8(2): 1-12.
Loomis G.A., Ries, S., Saywell, R.M. and Thakker, N.R. (2002). If electronic medical records are so great, why aren't family physicians using them? The Journal of Family Practice 51: 636-41.
Marasovic, C., Kenney, C., Elliott, D. and Sindhusake, D. (1997). Attitudes of Australian nurses toward the implementation of a clinical information system. Computers in Nursing 15: 91-8.
Ministry of Health. (2004). November Report. Department of Planning and Followup, Health Ministry Services, Kuwait.
Murff, H.J. and Kannry, J. (2001). Physician satisfaction with two order entry systems. Journal of the American Medical Informatics Association 8(5): 499-509.
Otieno, G.O., Hinako, T., Motohiro, A., Daisuke, K. and Keiko, N. (2008). Measuring effectiveness of electronic medical records systems: towards building a composite index for benchmarking hospitals. International Journal of Medical Informatics 77(10): 657-669.
PACI. (2007). Directory of population and labor force. Public Authority of Civil Information, Kuwait.
Poissant, L., Pereira, J., Tamblyn, R. and Kawasumi, Y. (2005). The impact of electronic health records on time efficiency of physicians and nurses: a systematic review. Journal of the American Medical Informatics Association 12(5): 505-16.
Sharda, P., Das, A.K., Cohen, T.A. and Patel, V. (2006). Customizing clinical narratives for the electronic medical record interface using cognitive methods. International Journal of Medical Informatics 75: 346-368.
Simpson, G. and Kenrick, M. (1997). Nurses' attitudes toward computerization in clinical practice in a British general hospital. Computers in Nursing 15(1): 37-42.
Sittig, D.F., Kuperman, G.J. and Fiskio, J. (1999). Evaluating physician satisfaction regarding user interactions with an electronic medical record system. Proceedings of the AMIA Symposium:. 400-404. Bethesda, Maryland USA American Medical Informatics Association
Southon, G., Sauer, C. and Dampney, K. (1999). Lessons from a failed information systems initiative: issues for complex organizations. International Journal of Medical Informatics 55(1): 33-46.
Van der Meijden, M.J., Tange, H., Troost, J. and Hasman, A. (2001). Development and implementation of an EPR: how to encourage the user. International Journal of Medical Informatics 64(2-3): 173-85.
Van Der Meijden, M.J., Tange, H.J., Troost, J. and Hasman, A. (2003). Determinants of success of inpatient clinical information systems: a literature review. Journal of the American Medical Informatics Association 10(3): 235-43.
Wang, S.J., Middleton, B., Prosser, L.A., Bardon, C.G., Spurr, C.D., Carchidi, P.J., Kittler, A.F., Goldszer, R.C., Fairchild, D.G., Sussman, A.J., Kuperman, G.J. and Bates, D.W. (2003). A cost-benefit analysis of electronic medical records in primary care. The American Journal of Medicine 114(5): 397-403.
Yoo, S.K., Kim, D.K., Kim, J.C., Part, Y.J. and Chang, B.C. (2008). Implementation of a large scale hospital information infrastructure for multi-unit health-care services. Journal of Telemedicine and Telecare 14(3): 164-6.
Saadoun Faris Al-Azmi PhD
Department of Medical Records
College of Health Sciences
The Public Authority for Applied Education and Training,
State of Kuwait
Naser Al-Enezi PhD
Faculty of Allied Health Sciences
Corresponding author: Rafiq I Chowdhury MSc
Department of Health Information Administration
Faculty of Allied Health Sciences
P.O. Box 31470
Tel: 965 9503456
Fax: 965 4830937
email: rafi firstname.lastname@example.org
Table 1: Overall user reaction regarding the various aspects of EMR system CHARACTERISTICS TOTAL N % Is the system rigid? Rigid 57 17.2 Flexible 275 82.8 Is the system adequate? Inadequate 126 38.0 Adequate 206 62.0 Is the system diffi cult? Difficult 37 11.1 Easy 295 88.9 Is the system dull? Dull 167 50.3 Stimulating 165 49.7 Is the system frustrating? Frustrating 63 19.0 Satisfying 269 81.0 Is the system terrible? Very bad 121 36.4 Wonderful 211 63.6 NOTE: N may not be added to the total sample size due to missing values. Table 2: Distribution of Background characteristics of the respondents by overall user reaction CHARACTERISTICS OVERALL UREACTION SCORE (%) 0 1 2 3+ Age (in years) * [less than or equal to] 25 15.5 24.8 21.1 15.5 26-35 39.7 43.6 35.5 39.7 [greater than or equal to] 36 44.8 31.6 43.4 44.8 Gender Male 48.3 29.1 42.1 39.5 Female 51.7 70.9 57.9 60.5 Education Intermediate 36.2 29.1 26.3 23.5 Secondary 32.8 28.2 31.6 34.6 College or above 31.0 42.7 42.1 42.0 Nationality Kuwaiti 94.8 99.1 98.7 97.5 Non-Kuwaiti 5.2 .9 1.3 2.5 Duty shift Morning 94.8 99.1 98.7 97.5 Evening 5.2 .9 1.3 2.5 CHARACTERISTICS TOTAL N % Age (in years) * [less than or equal to] 25 76 22.9 26-35 144 43.4 [greater than or equal to] 36 112 33.7 Gender Male 126 38.0 Female 206 62.0 Education Intermediate 94 28.3 Secondary 104 31.3 College or above 134 40.4 Nationality Kuwaiti 325 97.9 Non-Kuwaiti 7 2.1 Duty shift Morning 228 68.7 Evening 104 31.3 NOTES: * Significant at 5% level. N may not be added to the total sample size due to missing values. Table 3: Distribution of computer related experience of the respondents by overall user reaction CHARACTERISTICS OVERALL UREACTION SCORE (%) 0 1 2 3+ Previous PC experience Yes 63.8 67.5 73.7 65.4 No 36.2 32.5 26.3 34.6 Computer at home Yes 65.5 68.4 77.6 70.4 No 34.5 31.6 22.4 29.6 Computer at work Yes 87.9 86.3 85.5 80.2 No 12.1 13.7 14.5 19.8 Daily PC use frequency No use 5.2 14.5 11.8 16.0 Moderate 41.4 41.0 52.6 34.6 Many times 53.4 44.4 35.5 49.4 Typing ability ** Bad 13.8 5.1 9.2 23.5 Moderate 46.6 33.3 36.8 43.2 Good 39.7 61.5 53.9 33.3 Ease of data input ** Easy 98.3 94.0 93.4 76.5 Diffi cult 1.7 6.0 6.6 23.5 Computer error frequency ** Seldom 44.8 57.3 48.7 34.6 Sometimes 44.8 37.6 51.3 63.0 Always 10.3 5.1 2.5 Can access medical information from the electronic record * Yes 27.6 29.1 15.8 12.3 No 72.4 70.9 84.2 87.7 TOTAL N % Previous PC experience Yes 225 67.8 No 107 32.2 Computer at home Yes 234 70.5 No 98 29.5 Computer at work Yes 282 84.9 No 50 15.1 Daily PC use frequency No use 42 12.7 Moderate 140 42.2 Many times 150 45.2 Typing ability ** Bad 40 12.0 Moderate 129 38.9 Good 163 49.1 Ease of data input ** Easy 300 90.4 Diffi cult 32 9.6 Computer error frequency ** Seldom 158 47.6 Sometimes 160 48.2 Always 14 4.2 Can access medical information from the electronic record * Yes 72 21.7 No 260 78.3 NOTES: ** Significant at 1% level; * Significant at 5% level. N may not be added to the total sample size due to missing values. Table 4: Results of Poission Regression Analysis (dependent variable = overall user reaction) VARIABLES COEFFICIENTS p-VALUE Age (in years) -0.010 0.128 Gender (Male=1, Female=0) 0.051 0.711 Education Intermediate (Reference category) -- -- Secondary 0.078 0.493 College or above 0.079 0.497 Nationality (Kuwaiti=1, Others=0) 0.170 0.623 Duty shift (Morning=1, Evening=0) -0.221 0.118 Previous PC experience (Yes=1, No=0) -0.040 0.699 PC at home (Yes=1, No=0) 0.116 0.262 PC at work (Yes=1, No=0) -0.173 0.129 Daily PC use (Yes=1, No=0) 0.023 0.789 Typing ability Bad (Reference category) -- -- Moderate -0.169 0.206 Good -0.219 0.118 Ease of data input (Yes=1, No=0) -0.630 0.000 Computer error frequency (Seldom/sometimes=0, always=1) 0.631 0.028 Constant 0.020 1.672 Model Chi-square 55.758 0.000