JMIR Publications

ABSTRACT

Background:

Virtual patients are interactive digital simulations of clinical scenarios for the purpose of health professions education. There is no current collated evidence of the effectiveness of this form of education

Objective:

Our goal was to evaluate the effectiveness of virtual patients compared with traditional education, blended with traditional education, compared with other types of digital education and design variants of virtual patients in health professions education. The outcomes of interest were knowledge, skills, attitudes and satisfaction.

Methods:

We performed a systematic review of the effectiveness of virtual patient simulations in pre and post-registration health professions education following Cochrane methodology. We searched seven databases from the year 1990 up to September 2018. No language restrictions were applied. We included randomised controlled trials and cluster-randomised trials. We independently selected studies, extracted data and assessed risk of bias, then compared the information in pairs. We contacted study authors for additional information if necessary. All pooled analyses were based on random-effects models.

Results:

Fifty-one trials involving 4696 participants met inclusion criteria. Twenty-five studies compared virtual patients with traditional education, 11 studies investigated virtual patients as blended learning, five studies compared virtual patients with different forms of digital education, and 10 studies compared different design variants. The pooled effect of virtual patients comparison to traditional education (SMD=0.11, 95% CI:[-0.17,0.39], I2=74%, n=927) for knowledge and (SMD=0.90, 95% CI:[0.49,1.32], I2=88%, n=897) for skills. Studies measuring attitudes and satisfaction predominantly used surveys with item-by-item comparison. Trials comparing virtual patients with different forms of digital education and design variants were not numerous enough to give clear recommendations. Several methodological limitations in the included studies and heterogeneity contribute to a generally low quality of evidence.

Conclusions:

Low to modest and mixed evidence suggests that when compared to traditional education virtual patients can more effectively improve skills, and at least as effectively improve knowledge. We found evidence of effectiveness in both high and low-and-middle income countries demonstrating the global applicability of virtual patients. Further research should explore the utility of different design variants of virtual patients.