Provider satisfaction with the health information system based on the electronic health records in Burundi’s hospitals

INTRODUCTION

In computer language, the health record represents all the features of the software that contain administrative and medical information (care, prescriptions, etc.) and all correspondence between health professionals. The electronic health record (EHR) ensures the traceability of all actions carried out by health professionals.^1,2 It is a tool for communication, coordination and information between providers on the one hand and between providers and patients on the other. It allows to follow up and understand the patient's hospital journey, it is a primary element of the care quality by allowing their continuity in the context of a multi-professional and multidisciplinary care.³

With the emergence of Information and Communication Technologies, African governments have adopted resource mobilization strategies to make the use of electronic health records effective in health facilities. Projects to implement and use the electronic health record are thus newsworthy topics; however, their success depends on the adherence and acceptance of stakeholders at different levels of the health system in general by providers in particular.^4-5

Although Information and Communication Technologies seem to be a vehicle for the coordination of health professionals, the optimization of health expenses through a good organization of care processes and close cooperation to allow better care of patients, projects for the computerization of health records have not always been successful.⁵ According to some authors, the success of the electronic health records implementation project can be assessed on several axes, including the dimension of quality, utilization and user satisfaction. However, regardless of the results of the quality assessments of the EHR-based system, if users are not satisfied, they may not want to use it. Then the resistance of the users was suspected to be the main factor in the failure of the implementation of the EHR projects.⁵

In order to strengthen the Hospital Information System, the Ministry of Public Health and the Fight against AIDS in Burundi has started the project to computerize the medical record in hospitals. In 2015, computerization began with a pilot phase in 4 hospitals representative of the 3 levels (the district level, the regional level and the national level) of the health pyramid. For this purpose, one national hospital, one regional hospital and two district hospitals were selected to be part of the first pilot phase of computerization. In the logic of making a gradual extension of the computerization of hospitals in 2017, seven hospitals (tree national hospitals, one regional hospital and tree district hospitals) were equipped and started to use the computerized patient record software.^6-7

The software that was chosen to computerize the patient record is Open Clinic GA. It is an open-source software and can be redistributed or modified, marketable versions are also available. The main modules of the Open Clinic GA software concern the patient's administrative record, the patient's financial record, the patient's medical record, health insurance, credit unions, pharmacy (including stock), laboratory, radiology, health statistics in the form of periodic reports, human resources, the clinical thesaurus with coding assistance validated for ICD-10, SNOMED and multimedia media (images, video, audio).⁸

The purpose of this research was to describe and analyze the factors associated with the satisfaction of providers with the use of the electronic patient record.

METHODS

Type and framework of study

This was a descriptive and analytical cross-sectional study. Eleven hospitals whose health records were computerized with Open Clinic GA software were included in the study. Five of them belonged to the third reference level (district), two to the second reference level (region) and four to the first reference level (national). Four hospitals were computerized in 2015 and seven computerized in 2017. They were all beneficiaries of a computerization project funded by an international organization with bilateral cooperation.

Study population and organization of data collection

To fully explore provider satisfaction, a questionnaire was developed and self-administered to caregivers and administrative staff. The sample size was calculated from 155 providers with the Raosoft software, a proportion allocation was made to determine the number of subjects to be investigated in each hospital. Simple random selection was used to identify providers to be included in the study at two per service. The data were collected in February 2020.

Data capture, processing and analysis

The collected data was entered with the EPI INFO software, the analysis was done with STATA 15. To investigate the factors that explain the satisfaction of providers with the new system based on the use of the electronic health record, the bivariate analysis was done using the Pearson Chi-2 statistical test. Associations between the dependent variable and the independent variables were measured by Odds Ratio (OR) and their confidence interval at 95%. A multi-varied analysis according to the logistic regression model was made to identify the independent variables associated with the satisfaction of the providers. All independent variables whose significance was less than or equal to 25% in the bivariate analysis were included in the initial model. Step-by-step top-down modeling was performed to determine statistically significant variables and at the end, a final model was selected. The significance threshold was 5%. The Hosmer-Lemeshow test was used to determine the suitability of the final model.

RESULTS

Description of the characteristics of the providers

Of the 155 recipients included in the study, the majority were 52.9% male. The median age was 36 years with the minimum age of 24 and the maximum age of 74. The majority of providers were nurses (51.0%) followed by physicians (22.5%) (Table 1).

Satisfaction of service providers

Of the 155 providers, 86 (55.5%) were satisfied with the new EHR-based system. Of the 86 providers satisfied with the new EHR-based system, the perceived benefits of computerizing medical records are good management of health records and information (25.6%); continuity and quality of care (22.1%); good management of resources (12.8%), reduced service time (9.3%) and easy exchange of information between providers (7.0%); 4.7% of providers also referred to the ease of the software in producing reports and statistics.

The problems expressed by those who were not satisfied are (i) the software malfunction (lack of some diagnostics, updating of information on the EHR, interoperability and report generation problems in some services) expressed by 28.6% of providers; the need for training (16.3%), the insecurity of computers and tablets (15.3%), network problems (14.3%); the low capacity of providers to use the tool computer (10.2%), electrical power outages (7.1%). Difficulties in completing the EHR on time in emergency departments were expressed by 4.1% of providers.

Providers' knowledge and practices in using the software

The majority of study participants reported having the ability to easily use the authorizing officer (92.9%) and the software (87.7%). The majority of providers were trained by project computer scientists and those from the Ministry of Public Health and AIDS Control (45.2%), but a significant proportion had unmet training needs (85.8%) (Table 2).

Factors associated with provider satisfaction with the EHR-based hospital information system

In this study, variables such as gender, the profile of the person who trained providers on the software at the beginning of implementation, and the use of the software to produce reports and statistics were positively associated with providers ' satisfaction with the EHR-based hospital information system. Men were 2.21 percent more satisfied with the use of EHR than women (OR=2.21; 95% CI: [1.09-4.46]). Providers trained by project's hospital computer were 4.24 more satisfied than those trained by their colleges (OR=4.24; 95% CI: [1.47-12.23]) and 3.69 more satisfied than those trained by project computer scientists (OR=3.69; 95% CI: [1.25-10.89]). Similarly, the use of software to output reports and statistics was 2.27 times associated with provider satisfaction (OR=2.27; 95% CI: [1.14-4.49].

DISCUSSION

The assessment of the level of provider’s satisfaction was made through the self-administered questionnaire. The frequency of providers who are satisfied and the factors underlying their satisfaction were determined. The multivariate logistic regression analysis identified the factors associated with provider satisfaction in the 11 study hospitals. Of the 15 explanatory variables, six had significance below 25% in the bi-variable analysis were introduced into the initial model (table 4). Step-by-step top-down modeling was performed to determine statistically significant explanatory variables and at the end, a final model was selected. The Hosmer-Lemshow test concluded that the final model was satisfactory because the resulting p-value was 0.7699.

Of the 155 providers surveyed, the majority of providers were nurses (51.0%) followed by physicians (22.5%). Other providers included laboratory technicians, computer scientists, administrative and other support staff. In most provider studies on the implementation of the electronic health record, nurses are the majority. A study conducted in a hospital in Ethiopia on evaluating the implementation of the electronic health record, the majority of participants were nurses (64%) followed by doctors (27%).⁹ Another study on the preparation of health professionals to implement electronic health records system in three hospitals in Ethiopia showed that nurses were numerous in a proportion of 52.6%.¹⁰ This could be explained by the fact that in Africa, the availability of health personnel is estimated at 32 nurses per 10,000 inhabitants which is higher than the density of doctors per inhabitant (2 per 10,000 inhabitants).¹¹ Regarding satisfaction, the results of this study showed that 55.5% of providers were satisfied with the new EHR-based system, men were more satisfied than women (60.9% versus 49.3%). This proportion is lower than that found in a Ethiopia’s study where 65.6% were satisfied with the overall context of the EHR-based system.⁹

A study conducted in Saudi Arabia indicated that EHR user satisfaction was 40%.¹² The proportion of providers satisfied by the use of EHR differs from study to study. The totally different contexts of implementation of projects in terms of the state of the infrastructures and the characteristics of the providers could explain this difference.^13-14 The multivariate analysis by logistic regression showed that the gender, the exploitation of reports and statistics released in the software and the person who gave the training at the beginning of the computerization project were the three factors associated with the satisfaction of the providers in the 11 study hospitals. Compared to the person who gave the training, those who were trained by computer scientists specialists from the Ministry of Health and project level (62.9%) were more satisfied than those who were trained by computer scientists from the hospital (57.4%) and by colleagues (29.2%); the difference was statistically significant. This difference can be explained by the fact that the computer scientists `from the project had a higher skill gradient than other staff (colleagues or computer scientists from the hospital). The proportion of providers who used the software to produce reports or statistics was higher than those who said they had never released reports or statistics; 65.8% versus 44.7%. The management of reports or statistics in the paper-based system is more difficult and tedious, so it is understandable that providers who from time-to-time needed reports or statistics are more satisfied than those who claimed to have never released reports or statistics, it should also be noted that the computerization of the health record makes it possible to automate the production of reports and statistics.^2,3 Satisfaction with the EHR-based system was not associated with the demographic characteristics of participants, including age, qualification, or seniority in the hospital, which is consistent with several other studies.^12,15-17 This consistent finding between the studies reaffirms the importance of the performance (ease of use, interoperability etc.) of the software used to computerize the health record and the aspects related to the project implementation process regardless of the characteristics of the providers.

In this study the main limitation to be highlighted concerns the fact that it was carried out in public hospitals with the EHR system developed with the Open Clinic GA software, and the results may not be generalizable to other types of hospitals in particular those of private status or to hospitals with a different EHR system.

CONCLUSION

This study showed that 44.5% of providers were dissatisfied with the hospital information system based on EHR. Some of the reasons for the lack of satisfaction were, among others, the lack of updating of information on the EHR, problems of interoperability and generation of reports in some services, the need for training, insufficient computers, problems with networks; the low capacity of providers in the use of the computer tool and power cuts. The continuous evaluation of the computerization process and the observations given by users should guide stakeholders to undertake corrective actions to improve the adherence and satisfaction of providers to the hospital information system based on the EHR.

Funding: Project funded by European Union, Government of Burundi

Conflict of interest: None declared

Ethical approval: The study was approved by the National Ethics Committee for the protection of human subjects of biomedical and behavioral research

REFERENCES

1. 1. Sassi S, Verdier C. Presentation and visualization of medical documents. The digital medical record. Digital Document. 2009;12:37-58.
2. 2. Baujard C, Hamouda BI. Project management in the hospital: computerized patient record and quality of care. Res Manag Sci. 2015;109:147-64.
3. 3. Dupont D, Beresniak A, Kalra D. Interest of hospital electronic health records for clinical research. Med Sci. 2018;34:972-7.
4. 4. Bagayoko CO, Mueller H, Geissbuhler A. Assessment of Internet-based tele-medicine in Africa (the RAFT Project). Comput Med Imaging Graph. 2006;30:407-16.
5. 5. Makhloufi I, Saadi J, Artiba A. Analysis of risk factors related to the project of urbanization of the hospital information system of Cheikh Zaid Hospital. Revue. 2009;1:1-8.
6. 6. Ministry of Public Health and AIDS Control. Plan national de développement de l’informatique de santé du Burundi (PNDIS). An e-health enterprise architecture for the Ministry of Public Health and the Fight against AIDS of Burundi. Available at: https://www.researchgate.net/publication/342146340_Plan_National_de_Developpement_de_l'Informatique_de_Sante_du_Burundi_2020-2024_PNDIS_II. Accessed on 29 July 2019.
7. 7. Ministry of Public Health and AIDS Control. Strategic plan of the health information system. 2019-2023.
8. 8. Verbeke F, Gillaer D, Neto DJE. Open Clinic User manual. MXS SA for Eb@le Santé; 2010:98. Available at: https://fr.scribd.com/document/212952271/Openclinic-Manual-Fr-Ebale-Sante.
9. 9. Berhe M, Tadesse K, Berhe G. Evaluation of electronic medical record implementation from user’s perspectives in Ayder Referral Hospital Ethiopia. J Health Med Inform. 2017;8:1000249.
10. 10. Biruk S, Yilma T, Andualem M. Health Professionals’ readiness to implement electronic medical record system at three hospitals in Ethiopia: a cross sectional study. BMC Med Inform Decis Mak. 2014;14:115.
11. 11. World Health Organization. Global health statistics. Geneva: World Health Organization, 2009:95-105.
12. 12. Alharthi H, Youssef A, Radwan S. Physician satisfaction with electronic medical records in a major Saudi government hospital. J Taibah Univ Med Sci. 2014;9:213-8.
13. 13. Verbeke F, Karara G, Nyssen M. Human factors predicting failure and success in hospital information system implementations in Sub-Saharan Africa. Stud Health Technol Inform. 2015;216:482-6.
14. 14. Wubah AA, Yankson JA, Sumpter C. Evaluation of an ICD logging system to supplement an EMR in a Sub-Saharan country. 2018;7:8-13.
15. 15. Chisolm DJ, Purnell TS, Cohen DM. Clinician perceptions of an electronic medical record during the first year of implementation in emergency services. Pediatr Emerg Care. 2010;26:107-10.
16. 16. Weimar C. Electronic health care advances, physician frustration grows. Phys Exec. 2009;35:8-15.
17. 17. Morton ME, Wiedenbeck S. A framework for predicting HER adoption attitudes: a physician survey. Perspect Health Inf Manag. 2009;6:1a.