Objective To systematically catalogue review studies on digital health to establish extent of evidence on quality healthcare and illuminate gaps for new understanding, perspectives and insights for evidence-informed policies and practices. Methods We systematically searched PubMed database using sensitive search strings. Two reviewers independently conducted two-phase selection via title and abstract, followed by full-text appraisal. Consensuses were derived for any discrepancies. A standardized data extraction tool was used for reliable data mining. Results A total of 54 reviews from year 2014 to 2021 were included with notable increase in trend of publications. Systematic reviews constituted the majority (61.1%, (37.0% with meta-analyses)) followed by scoping reviews (38.9%). Domains of quality being reviewed include effectiveness (75.9%), accessibility (33.3%), patient safety (31.5%), efficiency (25.9%), patient-centred care (20.4%) and equity (16.7%). Mobile apps and computer-based were the commonest (79.6%) modalities. Strategies for effective intervention via digital health included engineering improved health behaviour (50.0%), better clinical assessment (35.1%), treatment compliance (33.3%) and enhanced coordination of care (24.1%). Psychiatry was the discipline with the most topics being reviewed for digital health (20.3%). Conclusion Digital health reviews reported findings that were skewed towards improving the effectiveness of intervention via mHealth applications, and predominantly related to mental health and behavioural therapies. There were considerable gaps on review of evidence on digital health for cost efficiency, equitable healthcare and patient-centred care. Future empirical and review studies may investigate the association between fields of practice and tendency to adopt and research the use of digital health to improve care.

No abstract

The application of artificial intelligence (AI) is on the rise in the healthcare industry. However, the study on the physicians’ perspectives is still lacking.  The study aimed to examine physicians’ attitudes, expectations, and concerns regarding the application of AI in medicine. A cross-sectional study was conducted in October 2019 among physicians in a tertiary teaching hospital in Malaysia. The survey used a validated questionnaire from the literature, which covered: (1) socio-demographic profile; (2) attitude towards the application of AI; (3) expected application in medicine; and (4) possible risks of using AI. Comparison of the mean score between the groups using a t-test or one-way analysis of variance (ANOVA). A total of 112 physicians participated in the study: 64.3% from the clinical departments; 35.7% from the non-clinical specialties. The physicians from non-clinical departments had significantly higher mean attitude score (mean = 14.94 ± 3.12) compared to the clinical (person-oriented) departments (mean = 14.13 ± 3.10) and clinical (technique-oriented) departments (mean = 13.06 ± 2.88) (p = 0.033). The tech-savvy participants had a significantly higher mean attitude score (mean = 14.72 ± 3.55) than the non–tech-savvy participants (mean = 13.21 ± 2.46) (p = 0.01). There are differences in the expectations among the respondents and some concerns exist especially on the legal aspect of AI application in medicine. Proper training and orientation should precede its implementation and must be appropriate to the physicians’ needs for its utilization and sustainability.

No abstract