Objective This review assesses interventions and their effectiveness in mitigating psychological consequences from pandemic.
Method Published English literatures were searched from four databases (Medline, PubMed, Embase and PsycINFO) from January 2020 and September 2021. A total of 27 papers with 29 studies (one paper reported three studies) met inclusion criteria. Cochrane risk-of-bias tool is applied to assess the quality of all randomised controlled trials (RCT).
Results All studies were recently conducted in 2020. Publications were from high-income (13, 44.8%), upper middle-income (12, 41.4%) and lower middle-income countries (3, 10.3%) and global (1, 3.5%). Half of the studies conducted for general population (51.7%). One-third of studies (8, 27.6%) provided interventions to patients with COVID-19 and 20.7% to healthcare workers. Of the 29 studies, 14 (48.3%) were RCT. All RCTs were assessed for risk of biases; five studies (15, 35.7%) had low risk as measured against all six dimensions reflecting high-quality study.
Of these 29 studies, 26 diagnostic or screening measures were applied; 8 (30.9%) for anxiety, 7 (26.9%) for depression, 5 (19.2%) for stress, 5 (19.2%) for insomnia and 1 (3.8%) for suicide. Measures used to assess the baseline and outcomes of interventions were standardised and widely applied by other studies with high level of reliability and validity. Of 11 RCT studies, 10 (90.9%) showed that anxiety interventions significantly lowered anxiety in intervention groups. Five of the six RCT studies (83.3%) had significantly reduced the level of depression. Most interventions for anxiety and stress were mindfulness and meditation based.
Conclusions Results from RCT studies (11%, 78.6%) were effective in mitigating psychological consequences from COVID-19 pandemic when applied to healthcare workers, patients with COVID-19 and general population. These effective interventions can be applied and scaled up in other country settings through adaptation of modes of delivery suitable to country resources, pandemic and health system context.
- mental health
- public health
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
Effective interventions which targeted patients, healthcare workers and general population can be applied and scaled up by other countries.
The heterogeneity of included studies such as the mode of delivery of interventions, participant characteristics and measurement tools prevents efforts to conduct meta-analysis.
There is a lack of assessment of the effects of cointerventions, provided to the participants in the intervention and control groups, which influence the outcomes in both groups.
Measurement of outcome at short duration after interventions does not offer understanding of long-term outcome or its sustainability.
COVID-19 was first reported in December 2019.1 By March 2020, the WHO declared the COVID-19 a pandemic and as of November 2021, there have been over 259 million cumulative confirmed cases and 5 million deaths from COVID-19 worldwide.2 The rapid global spread of disease has had physical health impacts and psychological consequences on the population.
The COVID-19 outbreak has had direct and indirect impacts on mental health. Many experienced fear and anxiety due to uncertainty about the pandemic’s evolution, feelings of hopelessness, despair and grief in the face of uncontrollable events.3 4 The impact of public health measures such as quarantine measures, social restrictions and isolation has also resulted in mental health consequences including anxiety, depression and feelings of neglect; anxiety was highly prevalent in people undergoing quarantine or social isolation.5–7 The pandemic’s indirect effects, such as unemployment due to economic downturn and the death of family members from disease, have led to depression and suicide.8–10 Decreased population mobility and high rates of COVID-19 infection were found to be significantly related to an increased prevalence of major depression disorder, which compared with prepandemic, there were additional 53.2 million cases of major depressive disorder and 76.2 million cases of anxiety disorders in 2020.11
Psychological consequences from the pandemic also developed more significantly in certain groups of people. Long hours of COVID-19 clinical services caused significant pandemic-related burnout for frontline healthcare workers worldwide.12–14 Fear of transmitting the infection to family members and increased demand for health services also caused stress and sleep problems among medical professionals.15–17
Multiple mental health interventions have been implemented during previous epidemics such as Ebola and Middle East respiratory syndrome, particularly for healthcare workers experiencing higher levels of burnout and psychological consequences than others.18 Effective interventions like psychosocial and administrative support in workplaces have been shown to reduce the rates of emotional exhaustion in healthcare workers.18 Interventions in other population groups during Ebola outbreaks reduced depression, stress and anxiety.19–21 Implementing interventions to alleviate mental health problems during the COVID-19 pandemic has proven to be difficult due to various social restrictions; face-to-face interventions seem not feasible where online interventions including telehealth and multimedia application were used as alternatives.22–24
Though various interventions to mitigate psychosocial consequences from the COVID-19 pandemic have been launched, the measures used, their delivery channels and effectiveness on population groups are not well understood. This systematic review assesses interventions and measures used, and the effectiveness of interventions in mitigating psychological consequences on different populations so that effective interventions can be upscaled accordingly. The psychological consequences affect patients who suffered from the illnesses and stress from quarantine and isolation, healthcare workers who suffered from stressful long hours of work and general population facing indirect effects of government measures such as lockdown and limited social activities.
We propose three review questions. First, what interventions are used to mitigate psychological consequences in the population and in healthcare workers? Second, what measures are used for assessing intervention outcome? Third, what is the effectiveness of different interventions? Findings from this review can inform health professionals and governments of effective interventions to prevent or mitigate mental health problems from the COVID-19 pandemic.
We used PICO framework, a common tool used to develop literature search strategies, to ensure the comprehensive searches recommended by the Cochrane Collaboration.25 It is used in evidence-based practice to formulate healthcare-related question. PICO elements include Population, Interventions, Comparisons and Outcomes. In this review, population are individuals in countries or areas affected by COVID-19; interventions are exposure to COVID-19 pandemic and the clinical or non-clinical interventions which mitigate the psychological impact from the pandemic; comparison of outcomes between case and control groups or before and after interventions; and outcome are mental consequences including suicide, depression, anxiety, stress and insomnia/reduced sleep quality. See the full search strategy in online supplemental annex 1.
Published literatures were electronically retrieved from four databases including Medline, PubMed, Embase and PsycINFO. The eligibility criteria are English-language studies between January 2020 and September 2021 where interventions were provided with measurable outcomes. Table 1 provides inclusion and exclusion criteria.
Selection of publications
The search terms were identified based on search strategy and applied to search in title, abstract, keyword and full text. After searching, the duplicated studies or data were removed then the abstracts were reviewed; those were excluded for the following reasons with agreement in inclusion and exclusion criteria.
Figure 1 shows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow of this process.
In the process of abstract screening, three researchers (AL, PP and AC) independently reviewed the abstracts. Consensus was reached if two researchers agreed based on eligibility criteria. In case of disagreement, the third researcher reviewed and decided.
Assessment of the risk of bias
The Cochrane risk-of-bias tool25 was applied to assess the methodological quality of 14 randomised controlled trials (RCT) included by this systematic review. There are six domains of risk of bias: (1) random sequence generation (selection bias), (2) allocation concealment (selection bias), (3) blinding of participants, personnel and outcome assessors, (4) incomplete outcome data (attrition bias), (5) selective reporting (reporting bias), and (6) other bias.25 Three authors (AL, PP and AC) separately evaluated each domain and categorised the risk of bias into ‘low risk’, ‘high risk’ or ‘unclear risk’. The results of assessment were shown in terms of the number of ‘low risk’ of bias, which is 6 in total. Different criteria were used for making risk of bias judgements for each domain. Insufficient information reported in studies resulted in ‘unclear risk’ of bias judgement. When there were conflicting views, discussion among the three authors was convened to reach consensus.
Data extraction and synthesis
Relevant contents in all included articles were extracted and synthesised into three variables in an Excel spreadsheet: (a) characteristics of study: author, year of publication, objective, country, study design; (b) interventions: provider, participant; and (c) outcomes: different types of psychological consequence, and measures used for assessment of psychological consequence.
Patient and public involvement
No patients were involved in this study.
Our search strategies identified a total of 5623 records, of which 5619 were from the four databases and four from hand-searching from references of key articles. After abstract screening, 705 duplicated records were removed while 4602 records were not relevant. The full texts of the remaining 316 records were retrieved and reviewed for eligibility. We found that 289 publications were not relevant to the review questions. Finally, 27 articles comprising 29 studies were included for systematic review.
Characteristics of 27 included articles are described in table 2. One paper reported three studies,26 totalling 29 studies, all of which were conducted in 2020. Of the 29 studies, one was conducted globally (3.5%)27 and the rest were categorised according to WHO geographical area. Three studies (10.3%) were conducted in the Eastern Mediterranean region,28–30 10 (34.5%) in the European region,31–40 4 (13.8%) in the Americas41–44 and 11 (37.9%) in the Western Pacific.26 45–52 Applying the World Bank’s income group categories, 13 (44.8%) studies were in high-income group,31–33 35–41 43–45 12 (41.4%) in upper middle-income group26 34 42 46–52 and 3 (10.3%) in lower middle-income group28–30; there were no studies from the low-income group.
On population, out of the total 29 studies, 8 (27.6%) targeted patients with COVID-19,28–30 45 48 49 51 52 6 (20.7%) among healthcare workers31 34 35 44 46 47 and more than half (51.7%) were general population.26 27 32 33 36–43 50 Out of the 29 studies, 14 (48.3%) were RCTs.27–29 33 34 38–42 44(2B,3)48 50 The remaining 15 (51.7%) were non-RCT studies.
The interventions reported by the studies can be classified into three types: 19 online-based interventions (65.5%),26 27 29–31 33 35 36 38 40–45 47 48 6 on-site-based interventions (20.7%)28 34 39 49–51 and 4 combined online and on-site interventions (13.8%).32 37 46 52
On outcome measurement, out of the total 29 studies, only 1 (3.5%) assessed suicide ideation,45 while 14 (48.3%) measured depression outcome,32 35 38–41 43–47 50–52 24 (82.8%) measured anxiety,26 28 29 32 34–36 38–52 9 (31.0%) measured psychological stress,30 35–40 43 50 9 (31.0%) measured sleep quality27 33 35 43 45 47–49 52 and 6 (20.7%) measured other outcomes such as loneliness and self-compassion.26 31 36 41 44 Table 2 shows the characteristics of the studies.
Measures used for the assessment of psychological consequences
Out of 29 studies, 20 measures were used to assess psychological outcomes. We categorised the measures based on the outcome they measured including suicidal ideation depression, anxiety, stress and insomnia (online supplemental annex 2).
Suicidal ideation was assessed by using one measure, the 9-item Beck Depression Inventory (BDI), whereas depression and anxiety were assessed by using seven and eight measures, respectively. The Hospital Anxiety and Depression Scale (HADS) was most used for the assessment of depression, while the Spielberger State-Trait Anxiety Inventory (STAI) was used in six studies (25.0%) on anxiety.26 28 29 34 36 48 Two different measures were used to assess both depression and anxiety, including the Patient Health Questionnaire-4 and HADS. The Brief Symptom Inventory-18 and Depression Anxiety Stress Scale-21 were used to assess depression, anxiety and psychological distress.
Stress was evaluated primarily using five measures (55.6%) in nine studies, though the 10-item Perceived Stress Scale (PSS-10) was most commonly used. Among the other outcomes, insomnia assessments were distinctive in that they consisted of five tests, each of which measured the patients’ sleep quality. The Insomnia Severity Index was widely used to determine sleep quality in four studies.27 33 35 45 There were several measures assessing outcomes beyond the focus of this study, one test for self-compassion and two tests for loneliness.26 36 41 44
Interventions, measurements and effectiveness
Table 3 (RCT studies) and table 4 (non-RCT studies) describe study design, providers of intervention, target population, intervention, measurement and psychological outcome (see full description in online supplemental annex 3).
Of 24 anxiety interventions from 18 publications, 21 (87.5%) significantly lowered anxiety in intervention groups; 11 were RCT26 28 29 34 39 41–43 48 50 and 10 quasiexperimental studies.26 32 36 44–47 49 51 52 Among the 11 RCT studies, five interventions (45.5%) were mindfulness and meditation based.26 39 42 43 50 Two studies (18.2%) applied specific techniques such as emotional freedom techniques34 and progressive muscle relaxation and deep breathing techniques.48
Fourteen studies provided different interventions to reduce depression, of which 11 (78.6%) significantly reduced depression.32 38–41 43 44 46 50–52 Of 14 studies, 5 (35.7%) of them were RCT; 4 (28.6%) related to mindfulness and meditation-based stress reduction,43 50 focusing on mind-body exercises including yoga40 and tai chi,39 and 1 (7.1%) provided empathetic conversations through phone calls.41
Of nine studies on insomnia, seven (77.8%) found that the interventions significantly reduced insomnia or improved sleep quality.27 33 43 47–49 52 Of these nine studies, only two (22.2%) were RCT, with one (11.1%) using a mobile meditation application,43 and the other using progressive muscle relaxation and deep breathing techniques.48
Of the total nine interventions against stress, seven (77.8%) had shown significant positive outcomes in stress reduction.30 36–39 43 50 One study (11.1%) shows no association of interventions and outcome,40 while one (11.1%) study showed positive and negative results due to timing of assessment.45 Of the total 29 studies, three (10.3%) reported the effectiveness of interventions on self-compassion; participants reported significantly higher perception of self-compassion than those in the control group.26 36 44 One of 29 studies (3.4%) using preassessment and postassessment shows improvement of loneliness.41 Only one study (3.4%) examined suicide, 30 min telephone consultation was provided which was effective after 1 week.45
Risk of bias
The quality of reporting for 14 RCT studies was assessed using the Cochrane risk-of-bias tool. Table 5 shows the risk of bias by six domains. These RCT studies had no risk of attrition bias (incomplete outcome data) or reporting bias (selective reporting). Of 14 studies, five (35.7%) had low risk of bias for all six dimensions reflecting high-quality RCT. Most studies had unclear reporting on the risk of allocation concealment, random sequence generation and blinding participants and personnel.26 28 39 42 48 Two studies (14.3%) reported high risk of bias on blinding of participants and personnel.40 50
All 29 studies in this review were carried out in 2020 in all geographical continents except Africa and Australia. The majority of studies (25, 86.2%) were from high-income and upper middle-income countries. Therefore, the varieties of interventions in this review can be useful for adaptive application in different country contexts, although there is no study from low-income countries. Psychological impacts are caused by direct consequences of the pandemic,53–55 and by public health and social containment measures used by governments, notably movement restrictions and lockdowns.56 Over a half (19, 65.5%) of the psychological interventions reviewed in this study were delivered to patients with COVID-19, healthcare workers and general population via an online platform, as on-site intervention was not possible in the pandemic context. The psychological consequences from this pandemic are diverse, requiring a wide range of interventions and tools for measuring subsequent psychological outcomes, including suicidal ideation, depression, anxiety, stress and insomnia.
Each outcome was evaluated using a variety of tools. The BDI was the most commonly used to assess suicidal ideation, followed by the HADS for screening depression, the STAI for screening anxiety, the PSS for screening stress and the Insomnia Severity Index for screening insomnia. The measures were all used for screening rather than diagnosis. Several measures were validated, with Cronbach’s alpha values greater or equal to 0.7 indicating high reliability. The majority of the tools were self-assessment questionnaires that participants in the study could complete using the sample scoring system. However, no articles reported on the sensitivity and specificity of each measure.
Interventions and psychological outcomes
Overall, most interventions have positive impact in mitigating psychological consequences in patients with COVID-19, healthcare workers and general population. A single intervention can have combined effects on multiple psychological outcomes; for example, one RCT using mobile meditation application, including verbal and written encouragement, mindfulness meditation, sleep stories and nature sounds, has significant positive effects on reduction of depression, anxiety, stress and insomnia.43 However, another study using an online application targeting emotional skills, healthy lifestyle behaviour, burnout and social support reported no significant difference in the same measured outcomes.35
Almost all 14 RCT studies that included mindfulness and meditation interventions to reduce depression, anxiety and stress showed significantly improved outcomes (depression,39–41 43 50 anxiety,26 28 29 34 39–43 50 stress39 43 50) with the exception of one study,40 which did not improve anxiety. Stress, anxiety and depression are caused by sympathetic nervous system activation due to widespread depolarisation throughout the brain and body; on the other hand, meditation and deep breathing lead to parasympathetic nervous system activation due to widespread inhibition and hyperpolarisation.57 Recently, mindfulness and meditation-based interventions have increasingly been integrated into mental health interventions58; meditation practices facilitate training of the self-regulation of attention and awareness, thereby enhancing control of mental processes, and consequently increasing well-being. Meditation and breathing techniques such as yoga and tai chi, targeting body and mind, are also applied and are effective in reducing stress, anxiety and depression.59 60 Yoga, for example, helps with intellectual and mental exercise, improves the psychological management and monitoring of stress and negative emotions and supports mental balance.60–62 Mindfulness and meditation-based interventions are readily available, and do not pose threats or risks. It can be applied as a first-line or complementary intervention for stress, anxiety, depression and some emotional disorders during COVID-19.
Moreover, combining mindfulness intervention with online application technology can benefit people who are quarantined or in isolation wards with no direct access to psychotherapy. In this review, online multimedia psychoeducational interventions were demonstrated to significantly reduce stress30; mindfulness-based stress reduction, cognitive–behavioural techniques, stress management techniques and positive psychotherapy have been delivered via WhatsApp. Nowadays, internet-based, self-administrated multimedia and education can promote access to mental health interventions, particularly amidst COVID-19. Online-based interventions enable people to social distance and be time efficient and cost-effective.
The pandemic does not allow in-person on-site intervention, causing a major shift to remote care63–65 and telemedicine; the use of electronic communication to deliver healthcare services at a distance becomes a key supporting measure for health service delivery systems during this pandemic.66 67 Several channels of telemedicine are accessible by patients and health professionals including telephone calls, video calls, interactive mobile health applications (m-Health), short message service, email, secure remote patient monitoring and videoconferencing. Video call via social media is the most common channel of communication, but special communication software (eg, Zoom) or current messenger applications (eg, WhatsApp and WeChat) are popular and convenient options as well.68
Furthermore, several countries have developed digital health technology to facilitate healthcare delivery through an online platform.69 70 Several challenges remain; online interventions cannot replace in-person therapy, notably when there is high risk of suicide which necessitates the presence of a psychotherapist or human interaction.71 72 Furthermore, several online counsellors65 raised concerns from their experiences, such as a lack of non-verbal language for interpreting patients who struggle to verbally communicate their feelings73 or significantly shorter counselling periods through the telephone due to ‘more superficial responses to questions’.74 However, the pandemic has raised awareness for the necessity of the online interventions where limitations can be resolved, as several studies have reported benefits of online services.75–78
The most common reason across items of bias assessment was related to incomplete or unclear reporting in the RCT.79 Between 2011 and 2014, 24.7% of the RCTs included in the Cochrane Review demonstrated an unclear risk of bias in the domain of incomplete outcome data.80 But all 14 RCTs in the review presented complete outcome data and reported loss to follow-up for risk of attrition bias. In this review, each article was assessed by at least two authors and we compared the risk of bias assessment for each item in terms of ‘high’, ‘low’ or ‘unclear’ risk of bias between the two reviews.
More than half of the studies had unclear risk on allocation concealment because the method of concealment was not described, or due to inadequate detail to allow a clear judgement. Similarly, 50.1% of the reported April 2011 issue of the Cochrane Database of Systematic Reviews had unclear risk of bias on allocation concealment.81 Concealment of randomised allocation prevents an influence of patient characteristics on allocation to intervention and control groups. One study found that lack of adequate allocation concealment was associated with overestimation of treatment effect.82
Two studies reported high risk of bias on blinding of participants and personnel.40 50 Blinding of participants and personnel prevents differences in patient management between groups and blinding of outcome assessors also prevents knowledge of the assigned intervention group influencing outcome measurement. However, both studies were conducted through online platform and the group memberships were not blind.
Causal interpretations are limited by 15 non-randomised controlled study designs. In addition to variation in sample size across studies, there were also variations in the mode of delivery of interventions, participant characteristics and measurement tools across different study sites. Furthermore, a majority of the trials included small sample sizes which is commonplace for studies on psychological interventions. High heterogeneity of selected studies led to inability to conduct meta-analysis. There was also a lack of assessment of cointerventions, which provide external sources of support to the participants in the intervention and control groups and may have influenced the results of the outcomes of an intervention. Lastly, postintervention outcomes from the studies were measured from short follow-up durations, so it is not possible to conclude the long-term psychological effects of those interventions.
Conclusion and policy implications
Of the total reviewed 29 studies, 26 diagnostic or screening measures were applied; 8 for anxiety (30.9%), 7 for depression (26.9%), 5 for stress (19.2%), 5 for insomnia (19.2%) and 1 for suicide (3.8%). Most studies conducted in high-income and upper middle-income countries. Half of studies targeted general population and the remaining in patients with COVID-19 and healthcare workers. Among 14 RCT studies, most interventions were effective in reducing psychological consequences.
Despite study limitations, the use of measurement tools and the strength of evidence from this review suggest an overall positive impact of the different interventions in terms of reducing psychological consequences in the population and healthcare workers. These effective interventions can be replicated and applied in other country settings, for which modes of delivery can be adjusted in line with country resources, epidemic situation and social and health system contexts. Moreover, it is necessary to improve the use of evidence in psychological interventions in mental health services.
We recommend that these effective interventions can be replicated and applied in other country settings, for which modes of delivery can be adjusted in line with country resources, epidemic situation and social and health system contexts. In the first quarter of 2022, the pandemic is still far from over due to low immunisation coverage in African countries and continued emergence of variants of concerns in the context of pressure to resuming the economic activities and opening the country borders.83 In ending the acute phase of the pandemic, rapid scale-up and promoting vaccine acceptance are key policy actions.84
Data availability statement
Data are available upon reasonable request.
Patient consent for publication
The authors gratefully acknowledge the funding support from the Thailand Science Research and Innovation (TSRI) under the Senior Research Scholar on Health Policy and System Research (contract number: RTA6280007) and the National Research Council of Thailand. The authors are grateful for the partial support from Chiang Mai University and wish to thank Saranya Sachdev for her support on proofreading.
Contributors Conceptualisation: AL, PP, AC and VT. Data extraction: AL, PP and AC. Formal analysis: AL, PP and AC. Writing—original draft preparation: AL, PP and AC. Writing—review and editing: CS and VT. All authors have read and agreed to the published version of the manuscript. The scientific guarantor of this publication is AL.
Funding This study is funded by the National Research Council of Thailand (NRCT) and the Senior Research Scholar on Health Policy and System Research (grant number: RTA6280007).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.