Fall prevention and management of behavioural and psychological symptoms of dementia (BPSD) in long-term care (LTC) facility is a major challenge. The objective of this systematic review is to assess the evidence of digital technology in their management. All studies of English-language excluding case-reports were eligible for review. Databases chosen were MEDLINE, EMBASE, Scopus, Web of Science and PSYCINFO from January 2000 to June 2020. Downs and Black checklist was used to check for risk of bias. Papers with a focus in LTC setting, using digital technology as intervention for older adults with dementia, and with measurable outcomes (outcomes that are quantified, not descriptive) were included in the final review. Seventeen original papers (8 RCTs, 8 quasi-experimental and 1 mixed method) were included. Three articles examining position-sensor technology for fall prevention showed mixed results. Two showed no difference and 1 showed small reduction in fall after alarm removal but the positive effect might be due to bias. Overall, the sample sizes were too small to draw meaningful conclusion. Fourteen studies (9 pet robots of which 8 were robotic seal/PARO) were identified for BPSD and results were mixed. Overall, PARO might have modest benefit in BPSD compared to usual care but might be no better than plush toy with more hallucinations or delusions seen in advanced dementia. However, the significant heterogeneity in methodology (intervention intensity, lack of record in psychoactive drug use), clinical tools used (different BPSD scales, different digital technologies) and variability in outcomes made it difficult to draw clear-cut conclusion. Studies involving other digital technologies are scarce and in pilot phases; hence, conclusion is premature. One limitation of the review was that only 9 out of 17 studies were of good quality. The limited research work in position-sensors meant insufficient evidence to prove efficacy for their use in LTC setting. The possible modest benefit of PARO in BPSD (e.g. in agitation, apathy or reduction in psychoactive drugs) was off-set by possible adverse events such as delusions or hallucinations in advanced dementia.

在长期护理 (LTC) 机构中，跌倒预防和痴呆症行为和心理症状 (BPSD) 的管理是一项重大挑战。本次系统审查的目的是评估数字技术在其管理中的证据。除病例报告外，所有英语研究均符合审查资格。选择的数据库为 2000 年 1 月至 2020 年 6 月的 MEDLINE、EMBASE、Scopus、Web of Science 和 PSYCINFO。Downs 和 Black 检查表用于检查偏倚风险。最终评审中纳入了关注长期护理（LTC）环境、使用数字技术对患有痴呆症的老年人进行干预以及具有可衡量结果（量化而非描述性结果）的论文。纳入 17 篇原始论文（8 篇随机对照试验、8 篇准实验论文和 1 篇混合方法）。三篇研究用于预防跌倒的位置传感器技术的文章显示了不同的结果。警报消除后，其中 2 个没有显示差异，1 个显示跌倒次数略有减少，但积极效果可能是由于偏差造成的。总体而言，样本量太小，无法得出有意义的结论。针对 BPSD 确定了 14 项研究（9 项宠物机器人，其中 8 项是海豹机器人/PARO），结果好坏参半。总体而言，与常规护理相比，PARO 可能对 BPSD 有一定的益处，但可能并不比毛绒玩具好，在晚期痴呆症中会出现更多的幻觉或妄想。然而，方法论（干预强度、缺乏精神活性药物使用记录）、使用的临床工具（不同的 BPSD 量表、不同的数字技术）和结果的差异性显着异质性，使得很难得出明确的结论。涉及其他数字技术的研究很少且处于试点阶段；因此，下结论还为时过早。该审查的一个局限性是 17 项研究中只有 9 项质量良好。位置传感器的研究工作有限，这意味着没有足够的证据来证明其在长期护理环境中的使用功效。PARO 在 BPSD 中可能带来的适度益处（例如，激动、冷漠或减少精神活性药物）被可能的不良事件（例如晚期痴呆中的妄想或幻觉）所抵消。