Applying Theoretical Domains Framework to investigate barriers and facilitators implementing an information system for elder’s post-discharge self-care: an exploratory sequential mixed-methods study (Preprint)

Abstract

Elderly patients face risks upon discharge due to limited health literacy, leading to adverse events and burden on the healthcare system. Digital health technologies have shown promise in improving discharge processes and patient-provider communication. Hong Kong government developed an information system to enhance patient-centered care by providing medication information and follow-up appointments. However, concerns exist regarding the fidelity of PDIS implementation.

This study aims to explore determinants impacting PDIS implementation and their association with implementation frequency with the guidance of a comprehensive implementation framework.

This study used an exploratory sequential mixed-method design. In stage one, semi-structured interviews with 53 purposively selected nurses informed the exploration of factors influencing the distribution and explanation of the PDIS form with direct content analysis. In stage two, a cross-sectional survey, informed by the qualitative findings, was conducted with 482 nurses to identify the most relevant and important determinants with generalized linear regressions. The qualitative and quantitative results were combined to generate meta-inferences and a joint display of the final results.

Qualitative interviews identified 36 statements and 13 domains as major factors, including PDIS characteristics, knowledge, attitudes, role agreement, self-efficacy, goal-setting, information circulation, and communication among nurses. Quantitative findings showed that 70% of nurses provided printed PDIS summaries for at least 90% of patients, and only 34% consistently explained it to at least 90% of patients. Regression analysis identified nine domains as relevant and important for both distribution and explanation, including knowledge (AOR: 1.74, P<0.001 for distribution frequency; AOR: 2.43, P <0.001 for explanation frequency), skills (AOR: 2.17, P <0.001 for explanation frequency), role identity (AOR: 1.74, P =0.02 for distribution frequency; AOR: 2.43, P <0.001 for explanation frequency), beliefs in consequences (AOR: 1.66, P <0.001 for distribution frequency; AOR: 1.68, P <0.001 for explanation frequency), beliefs in capabilities (AOR: 1.65, P <0.001 for distribution frequency; AOR: 2.05, P <0.001 for explanation frequency), intentions (AOR: 1.66, P <0.001 for distribution frequency; AOR: 1.61, P <0.001 for explanation frequency), goals (AOR: 1.29, P =0.02 for distribution frequency), memory and decision processes (AOR: 2.72, P <0.001 for distribution frequency; AOR: 1.89, P <0.001 for explanation frequency), and environmental context (AOR: 3.05, P <0.001 for distribution frequency; AOR: 2.44, P <0.001 for explanation frequency).

Our findings confirmed previous evidence on determinants, including knowledge, competencies, perceived effectiveness, role agreement, intentions, decision processes, and environmental context. Additionally, we highlighted the importance of goal setting for successful implementation. These results can guide practitioners in implementing similar initiatives and support evidence-based decision-making. Researchers can also further investigate the relationships between the identified determinants.

Not Applicable