Yao Li, Ding Nannan, Yang Liping, Zhang Zhigang, Jiang Lingjie, Jiang Biantong, Wu Yuchen, Zhang Caiyun, Tian Jinhui
Department of Nursing, the First Hospital of Lanzhou University, Lanzhou 730000, Gansu, China.
Department of Intensive Care Unit, the First Hospital of Lanzhou University, Lanzhou 730000, Gansu, China.
Zhonghua Wei Zhong Bing Ji Jiu Yi Xue. 2020 Mar;32(3):350-356. doi: 10.3760/cma.j.cn121430-20200213-00180.
To investigate the cognitive impairment after intensive care unit (ICU) discharge and provide theoretical basis for prevention and intervention.
Studies about cognitive impairment after ICU discharge were retrieved in PubMed, Embase, Cochrane Library, Web of Science, Wanfang data, CNKI and SinoMed from their foundation to December 2019. The literature screening and data extraction were performed by two researchers independently, and the quality of different types of researches was evaluated using Cochrane Handbook 5.1.0, Newcastle-Ottawa scale (NOS) and agency for healthcare research and quality criteria (AHRQ). The Meta-analysis was performed by Stata 13.0 software. Sensitivity analysis was used to determine the reliability of the combined effect values. Funnel plot and Egger test were used to analyze publication bias. The non-parametric clipping was used to evaluate the impact of publication bias on the results.
A total of 35 studies were enrolled, including 27 prospective cohort studies, 4 retrospective cohort studies, 2 randomized controlled trial (RCT) studies, 1 case-control study, and 1 cross-sectional study. Three literatures were published in Chinese and 32 were in English, which covered 13 countries, and a total of 102 504 ICU survivors were followed up successfully. Literature quality evaluation results showed that the NOS scores of 31 cohort studies were between 6 and 9, of which the case-control study scored 9. The quality grade of 2 RCT studies were both B. According to the AHRQ criteria, 1 cross-sectional study's design was scientifically rigorous and of high quality. Thirty-five studies reported that the overall incidence of cognitive impairment after ICU discharge ranged from 2.47% to 66.07%. For the multiple follow-ups studies, the first survey data was selected for Meta-analysis, and the results showed that the pooled incidence was 38.44% [95% confidence interval (95%CI) was 29.32-47.55]. Each study was removed for sensitivity analysis and the pooled results did not change much, which indicated that the results were reliable. The sub-group analysis was performed on different evaluation methods for cognitive impairment after ICU discharge, different types of ICU patients, and different follow-up time. The results showed that the pooled incidence of studies using neuropsychological test to evaluate cognitive impairment after ICU discharge was 31.42% (95%CI was 21.82-41.02), the pooled incidence of studies using questionnaires or scales was 38.75% (95%CI was 29.54-47.96), and the difference between the two groups was statistically significant (P < 0.01). The pooled incidence of cognitive impairment after ICU discharge in general ICU patients was 43.42% (95%CI was 30.88-55.95), acute respiratory distress syndrome (ARDS) patients' pooled incidence was 34.40% (95%CI was 23.02-45.79), and the pooled incidence of elderly ICU patients was 12.93% (95%CI was 8.48-17.37), the difference among the three groups was statistically significant (P < 0.01). The incidences of cognitive impairment < 1 year, 1 to 4 years, ≥ 5 years after ICU discharge were 43.30% (95%CI was 29.47-57.13), 34.21% (95%CI was 26.70-41.72), and 20.22% (95%CI was 4.89-35.55), respectively, and the differences among the three groups were statistically significant (P < 0.01). The funnel plot showed that the distribution of all studies was asymmetric, and the Egger test result also suggested that there might be publication bias (P < 0.05). The non-parametric clipping was used to estimate the impact of publication bias on the results, and the result showed that the difference in the incidence of cognitive impairment after ICU discharge before and after non-parametric clipping was large, suggesting that publication bias might influence the stability of the research results.
The incidence of cognitive impairment after ICU discharge is relatively high and persistent for a long time, but diagnostic criteria of cognitive impairment and follow-up time are quite different. It is necessary to develop consistent evaluation criteria and rigorous designed research in the further.
探讨重症监护病房（ICU）出院后的认知障碍情况，为预防和干预提供理论依据。
检索PubMed、Embase、Cochrane图书馆、Web of Science、万方数据、中国知网和中国生物医学文献数据库中自建库至2019年12月关于ICU出院后认知障碍的研究。由两名研究人员独立进行文献筛选和数据提取，并采用Cochrane手册5.1.0、纽卡斯尔-渥太华量表（NOS）和医疗保健研究与质量机构标准（AHRQ）对不同类型研究的质量进行评价。使用Stata 13.0软件进行Meta分析。采用敏感性分析确定合并效应值的可靠性。采用漏斗图和Egger检验分析发表偏倚。采用非参数剪补法评估发表偏倚对结果的影响。
共纳入35项研究，包括27项前瞻性队列研究、4项回顾性队列研究、2项随机对照试验（RCT）研究、1项病例对照研究和1项横断面研究。3篇文献发表于中文，32篇发表于英文，涉及13个国家，共成功随访102 504例ICU幸存者。文献质量评价结果显示，31项队列研究的NOS评分在6至9分之间，其中病例对照研究评分为9分。2项RCT研究的质量等级均为B级。根据AHRQ标准，1项横断面研究的设计科学严谨，质量较高。35项研究报告ICU出院后认知障碍的总体发生率在2.47%至66.07%之间。对于多次随访研究，选择首次调查数据进行Meta分析，结果显示合并发生率为38.44%[95%置信区间（95%CI）为29.32 - 47.55]。每项研究进行敏感性分析后，合并结果变化不大，表明结果可靠。对ICU出院后认知障碍的不同评估方法、不同类型的ICU患者和不同随访时间进行亚组分析。结果显示，采用神经心理测试评估ICU出院后认知障碍的研究合并发生率为31.42%（95%CI为21.82 - 41.02），采用问卷或量表评估的研究合并发生率为38.75%（95%CI为29.54 - 47.96），两组差异有统计学意义（P < 0.01）。普通ICU患者ICU出院后认知障碍的合并发生率为43.42%（95%CI为30.88 - 55.95），急性呼吸窘迫综合征（ARDS）患者的合并发生率为34.40%（95%CI为23.02 - 45.79），老年ICU患者的合并发生率为12.93%（95%CI为8.48 - 17.37），三组差异有统计学意义（P < 0.01）。ICU出院后<1年、1至4年、≥5年认知障碍的发生率分别为43.30%（95%CI为29.47 - 57.13）、34.21%（95%CI为26.70 - 41.72）和20.22%（95%CI为4.89 - 35.55），三组差异有统计学意义（P < 0.01）。漏斗图显示所有研究的分布不对称，Egger检验结果也提示可能存在发表偏倚（P < 0.05）。采用非参数剪补法估计发表偏倚对结果的影响，结果显示非参数剪补前后ICU出院后认知障碍发生率差异较大，提示发表偏倚可能影响研究结果的稳定性。
ICU出院后认知障碍的发生率较高且持续时间较长，但认知障碍的诊断标准和随访时间差异较大。有必要进一步制定统一的评估标准并开展设计严谨的研究。
