Are we There Yet? Immersive Virtual Reality to Improve Cognitive Function in Dementia and Mild Cognitive Impairment

Meelad Sayma, BMBS, BSc; Remco Tuijt, BA, MSc; Claudia Cooper, BMBS, MRCPsych, MSc; Kate Walters, BMBS, BMedSci, MRCGP, MMed, MSc, PhD


Gerontologist. 2020;60(7):e502-e512. 

In This Article


Protocol and Registration

This systematic review of published literature was conducted according to PRISMA guidelines and the format of this report follows the PRISMA statement (Moher, Liberati, Tetzlaff, & Altman, 2009). Our review was registered with PROSPERO, registration number: CRD42019122295 (Chien, Khan, & Siassakos, 2012).

Inclusion and Exclusion Criteria

Studies were eligible for inclusion if they were primary research studies conducted on humans and focused on IVR studies, which evaluated IVR as an intervention for people with dementia or MCI. All study types were included, such as feasibility studies, pilot studies, and efficacy/effectiveness studies. Studies were excluded if the article was a conference abstract/proceeding (efforts were made to identify whether a full-paper version of the abstract identified had been published). Studies were excluded if they targeted physical rehabilitation only in patient groups that had Parkinson's disease (without documented cognitive impairment), and without analyzing outcomes of patients separated by their cognitive function. Studies were excluded if they solely looked at the general population of older people or those with subjective cognitive impairment, and did not report findings specifically for those with dementia or MCI. Published research was accessed if it was available between 1 January 2000 and 6 December 2018. Studies published before this date would not have had sufficient technology to produce an IVR experience (Cipresso, Giglioli, Raya, & Riva, 2018).

Search Strategy

Search terms were developed and based on other clinical IVR-based systematic reviews (Cano Porras, Siemonsma, Inzelberg, Zeilig, & Plotnik, 2018; Dockx et al., 2016). Search terms are described in Table 1; the strategy was developed utilizing the PICOS model (Methley, Campbell, Chew-Graham, McNally, & Cheraghi-Sohi, 2014). Five databases were searched: CINAHL, PsycINFO, PubMed/Medline, EMBase, and OVID Medline. Further material was sought by contacting the authors of initially included papers, hand-searching of key journals in the fields of IVR and dementia/MCI, and article reference lists. Key authors were also contacted to ask whether they had published any further article, forward citation tracking was also conducted to identify further articles of relevance that had cited our identified articles.

Study Selection

One author (M.S.) performed initial eligibility assessment by first checking the title and abstract to ensure only papers investigating virtual reality in dementia or MCI were included. Two authors (M.S. and R.T.) then assessed the full text of remaining articles to reach a consensus on articles for inclusion, based on the preselecting inclusion and exclusion criteria. Disagreements on included papers were discussed and resolved with K.W. and C.C.

Data Extraction and Risk of Bias Assessment

Data from included papers were manually extracted onto data collection tables. We collected data on participants (medical condition, sample size, age, time of diagnosis), type of intervention (method of IVR delivery, length of IVR intervention, number of sessions, setting of IVR delivery), reported outcomes, effectiveness of IVR, recruitment rates, and attrition rates. In addition, we extracted data on characteristics of the study (design, country, and setting). The included articles were assessed for bias using the Critical Skills Appraisal Programme (CASP) checklist for randomized control trials; however, due to the relatively new nature of the field, no studies were excluded based on their risk of bias. We also included an additional systemic bias assessment as part of the data extraction portion of study analyses.

Defining and Assessing Extent of Immersion

Immersion can be defined as the extent to which a user feels "present" in their VE (Slater & Wilbur, 1997). Immersion is difficult to quantify and is largely judged by the quality of the hardware and the software used in the intervention, for example, the use of higher definition displays may be considered to offer a more significant level of immersion (Bowman & McMahan, 2007; Rash, 2008; Slater & Wilbur, 1997). Presence, however, refers to the participant's sense of being within an environment created using technology, whereas when using the definition above, immersion is often defined by the range of technology used with the attempt of achieving presence (Cole, Crowle, Austwick, & Slater, 2009; Steuer, 2006). As a result, this article will also attempt to assess the quality of hardware used in each intervention and look to establish the extent to which presence was assessed in each study, to illustrate the level of immersion achieved in each intervention.

In addition, the concept of IVR has been misused frequently in the dementia and MCI literature to describe the use of any VE, even in a nonimmersive environment. For the purpose of our study, we regarded articles as utilizing IVR if the authors made an attempt to create an immersive experience using the items in Table 1, but not necessarily utilizing an HMD, as long as alternative attempts to visually immerse the subject were made.

We devised a simple scoring method to assess the level of immersion through the use and availability of hardware for the intervention (Table 2). As the experience of "presence" is more subjective, we were unable to objectively assess this but looked at each article to appraise the extent to which they sought to establish participant "presence." Haptic feedback, in this case, is the creation of a simulated touch experience, delivered through vibrations or other kinetic movements on skin (Okamura, 2009).

Theoretical Basis of Intervention

When developing novel medications, the mechanism of action is one of the first aspects that is considered and analyzed. However, it is widely acknowledged that clinical IVR literature rarely discusses theoretical aspects of mechanisms of action or how IVR theory applies to the problem being investigated (Garrett et al., 2018). We will also assess the extent to which theory on cognition and the interplay of IVR and the relevant intervention has been considered in developing each intervention.

Data Synthesis

We intended to perform a meta-analysis if there was sufficient suitable data, or otherwise conduct a narrative synthesis of the included studies.