Thai Hospitals' Adoption of Information Technology: A Theory Development and Nationwide Survey.

摘要

Background: With documented benefits and recent public policies, health information technology (IT) has received increasing attention in recent years. However, knowledge about Thailand's state of hospital IT adoption is lacking. The literature also identifies organizational management practices that are important to health IT implementation, but these factors are rarely included in quantitative analysis. Paucity of theoretical developments in the area also prevents a systematic approach to IT implementation.;Objective: To describe the current state of IT adoption in Thai hospitals and test a proposed model of organizational IT adoption that includes facilitating management practices and important hospital characteristics, motivated in part by Pare and Sicotte (2001)'s IT sophistication framework with modifications.;Materials and Methods: A nationwide mail survey was conducted using a developed instrument with established face and content validity in 1,302 hospitals in Thailand after a pilot study using five hospitals for pre-test purposes. Each hospital's IT chief or executive was asked to assess the degrees of specific technologies' adoption, IT-supported hospital functions, within- and outside-hospital information sharing, and presence of specific management practices, each in a 5-point Likert-type scale. Confirmatory and exploratory factor analyses were done, resulting in the rejection of the proposed model and a new set of IT adoption factors that fit the data better. Average scores for each of these new IT adoption aspects were analyzed descriptively to provide Thailand's baseline adoption levels. Construct and criterion validity was also assessed. Path analysis was used to test the proposed model of hospital IT adoption and identify associated organizational factors. Estimates for adoption of basic electronic health records (EHRs), comprehensive EHRs, and computerized physician order entry (CPOE) were also computed from relevant IT-supported functions for cross-study comparisons.;Results: The nationwide survey received a 70% response rate, but responding hospitals tended to be somewhat larger and public. Thai hospitals overall had acceptable levels of IT adoption, but information sharing outside the hospitals was very limited. When both outpatient and inpatient settings were considered, about 50% of responding hospitals had complete or partial basic EHR adoption and only 5% had comprehensive EHR adoption, but 90% had CPOE for medication orders. Adoption estimates for the outpatient setting alone were somewhat larger than the inpatient setting. Significant correlations among the different aspects of IT adoption and between these constructs and other criterion variables provide evidence for construct and criterion validity. In path analysis, after respecifying the model based on the factor patterns discovered from the data, the final model indicated significant effects of public status on adoption of infrastructural technologies such as networking and master patient index, as well as on internal information sharing. Bed size was positively associated with infrastructural technologies adoption but negatively associated with the levels of IT-supported clinical functions. Teaching status was not associated with any aspects of IT adoption in the path model. As hypothesized, the extent of facilitating operational IT management was associated with the levels of technology adoption and use of IT to support clinical EHR workflows (order entry and results viewing) as well as inpatient clinical documentation. These latter three constructs were also associated with the extent of internal information sharing, while the extent of external information sharing was associated with the levels of internal information sharing and IT support for inpatient clinical documentation.;Discussion: Thailand's adoption picture is very encouraging with many hospitals having some IT infrastructure in place, though adoption gaps still exist. The discovered IT adoption factors and the developed survey instrument had supporting evidence for its validity, and the final model resulting from path analysis provides a useful framework for health IT adoption in future IT adoption studies. The positive association between public status and IT adoption and lack of significant hypothesized association between IT adoption and bed size or teaching status were surprising but may reflect the unique health IT market and dynamics in Thailand.;Conclusion: Basic IT adoption in Thai hospitals appears to have passed the tipping point. Focus should be on adoption of more advanced technologies (such as comprehensive EHRs and clinical decision support systems) and ensuring that adoption translates into better processes and outcomes, as well as addressing barriers to health information exchange. The utility of the proposed framework is demonstrated, as is the importance of identified facilitating IT management practices. The final model from this study, named the Theory of Hospital Adoption of Information Systems (THAIS) here, should be cross-validated and refined in future studies.;Funding: This study was supported by a research grant from the Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand under Grant No. RD53065/year 2010.