Long-Term Consequences of COVID-19 at 6 Months and Above: A Systematic Review and Meta-Analysis

We aimed to review the data available to evaluate the long-term consequences of coronavirus disease 2019 (COVID-19) at 6 months and above. We searched relevant observational cohort studies up to 9 February 2022 in Pubmed, Embase, and Web of Science. Random-effects inverse-variance models were used to evaluate the Pooled Prevalence (PP) and its 95% confidence interval (CI) of long-term consequences. The Newcastle–Ottawa quality assessment scale was used to assess the quality of the included cohort studies. A total of 40 studies involving 10,945 cases of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection were included. Of the patients, 63.87% had at least one consequence at the 6 month follow-up, which decreased to 58.89% at 12 months. The most common symptoms were fatigue or muscle weakness (PP 6–12 m = 54.21%, PP ≥ 12 m = 34.22%) and mild dyspnea (Modified Medical Research Council Dyspnea Scale, mMRC = 0, PP 6–12 m = 74.60%, PP ≥ 12 m = 80.64%). Abnormal computerized tomography (CT; PP 6–12 m = 55.68%, PP ≥ 12 m = 43.76%) and lung diffuse function impairment, i.e., a carbon monoxide diffusing capacity (DLCO) of < 80% were common (PP 6–12 m = 49.10%, PP ≥ 12 m = 31.80%). Anxiety and depression (PP 6–12 m = 33.49%, PP ≥ 12 m = 35.40%) and pain or discomfort (PP 6–12 m = 33.26%, PP ≥ 12 m = 35.31%) were the most common problems that affected patients’ quality of life. Our findings suggest a significant long-term impact on health and quality of life due to COVID-19, and as waves of ASRS-CoV-2 infections emerge, the long-term effects of COVID-19 will not only increase the difficulty of care for COVID-19 survivors and the setting of public health policy but also might lead to another public health crisis following the current pandemic, which would also increase the global long-term burden of disease.


Introduction
Caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the pandemic of coronavirus disease 2019 (COVID-19) is currently still the greatest global public health challenge. Reported to the World Health Organization (WHO), globally, as of 5 April 2022, there have been more than 490 million confirmed cases of COVID-19, including more than 6 million deaths [1]. However, the natural history, clinical course, and long-term effects are still not fully understood [2]. While the majority of patients recover from COVID-19, for a significant number of people, the virus poses a range of serious long-term effects or complications, regardless if they are men or women, hospitalized or not, young or old, or even children [2,3]. On 6 October 2021, the WHO developed a clinical case definition of the post-COVID-19 condition by Delphi consensus: the post-COVID-19 condition occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). This study was registered on PROSPERO (CRD42022309720).
We included observational cohort studies that examined the long-term consequences of COVID-19 at 6 months and above. The following studies were excluded: (1) irrelevant to the subject of the meta-analysis, such as studies that did not use SARS-CoV-2 infection as the exposure; (2) insufficient data to calculate the prevalence of long-term COVID-19 consequences; (3) duplicate studies or overlapping participants; (4) reviews, editorials, conference papers, case series/reports, secondary analysis or animal experiments; (5) qualitative designs; and (6) studies that did not clarify the identification of COVID-19. For example, the confirmed diagnosis of COVID-19 via a reverse-transcription polymerase chain reaction (rt-PCR) test, serologic test, or other means was not mentioned in the text.
Studies were identified by two investigators (MYR and DJ) independently following the criteria above, while discrepancies were solved by consensus or with a third investigator (LQ).

Data Extraction
The following data were extracted from the selected studies: (1) basic information of the studies, including the first author, publication time, and country where the study was conducted; (2) characteristics of the study population, including the sample size, median age, gender, follow-up period, smoking status, severity of COVID-19, underlying diseases, admission to hospital or intensive care unit (ICU), and length of stay (LOS); (3) clinical features of COVID-19, including the number of cases with general COVID-19-related symptoms, respiratory symptoms, cardiovascular symptoms, gastrointestinal symptoms, and neurological symptoms, as well as the results of a pulmonary functional test (PFT) and chest computerized tomography (CT); (4) the number of cases with psychiatric problems; and (5) the number of cases with problems in 5 dimensions of the European Quality of Life Five-Dimension Five-Level Scale (EQ-5D-5L), which is an instrument developed for describing and valuing health-related quality of life by the EuroQol Group in 1987. A template was used for the primary data extraction, as shown in Supplementary Table S1.
The data extraction and determination of information eligibility were conducted by two investigators (MYR and DJ) independently following the criteria above, while discrepancies were solved by consensus or with a third investigator (LQ).

Quality Assessment and Risk of Bias
We used the Newcastle-Ottawa quality assessment scale to evaluate the risk of bias in the included cohort studies. Cohort studies were classified as having a low (≥7 stars), moderate (5-6 stars), or high risk of bias (≤4 stars), with an overall quality score of 9 stars. We used the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach to evaluate the evidence quality of the long-term consequences of COVID-19.
Quality assessment was conducted by two investigators (MYR and DJ) independently, while discrepancies were solved by consensus or with a third investigator (LQ).

Data Synthesis and Statistical Analysis
We performed a meta-analysis to estimate the Pooled Prevalence (PP) and its 95% confidence interval (CI) of the long-term consequences of COVID-19 at 6 months and above. We performed subgroup analyses by the follow-up period (6-12 months and ≥12 months), severity of COVID-19 (non-severe and severe; the non-severe group included mild and moderate COVID-19, and the severe group included severe and critical COVID-19), whether patients were hospitalized (inpatients and outpatients), and gender. Random-effects or fixed-effects models were used to pool the rates and adjusted estimates across studies separately, based on the heterogeneity among estimates (I 2 ). Fixed-effects models were used if I 2 ≤ 50%, which represents low to moderate heterogeneity, and random-effects models were used if I 2 ≥ 50%, representing substantial heterogeneity. The D-L method was used to estimate the tau square in the case of random-effects models. Publication bias was assessed by Harbord's modified test. All analyses were performed using Stata version 16.0 (Stata Corp, College Station, TX, USA).

Basic Characteristics
In the initial literature search, 5271 potential articles were identified up to 9 February 2022 (1459 in PubMed, 1894 in Embase, 1918 in Web of Science. A total of 2512 duplicates were excluded. After reading the titles and abstracts, 2566 articles were excluded based on the inclusion and exclusion criteria. Among the 193 studies under full-text review, 153 studies were excluded. Eventually, 40 studies were included in this meta-analysis based on the inclusion criteria . The literature retrieval flow chart is shown in Figure 1. across studies separately, based on the heterogeneity among estimates (I²). Fixed-effects models were used if I² ≤ 50%, which represents low to moderate heterogeneity, and random-effects models were used if I² ≥ 50%, representing substantial heterogeneity. The D-L method was used to estimate the tau square in the case of random-effects models. Publication bias was assessed by Harbord's modified test. All analyses were performed using Stata version 16.0 (Stata Corp, College Station, TX, USA).

Basic Characteristics
In the initial literature search, 5271 potential articles were identified up to 9 February 2022 (1459 in PubMed, 1894 in Embase, 1918 in Web of Science. A total of 2512 duplicates were excluded. After reading the titles and abstracts, 2566 articles were excluded based on the inclusion and exclusion criteria. Among the 193 studies under full-text review, 153 studies were excluded. Eventually, 40 studies were included in this meta-analysis based on the inclusion criteria . The literature retrieval flow chart is shown in Figure 1. The included studies were observational cohort studies describing the long-term consequences of COVID-19 at follow-up 6 months and above, which involved 10,945 cases of SARS-CoV-2 infection. A total of 26 studies described COVID-19 consequences at 6-12 months' follow-up and 19 studies described COVID-19 consequences at 12 months and above. The majority of the included studies were of great methodological rigor (i.e., 7-9 stars on the Newcastle-Ottawa Scale); only 2 included studies had 6 stars, mainly due to the insufficient comparability between the exposed cohort and unexposed cohort. The characteristics of the included studies are shown in Supplementary Table S2.

Pooled Prevalence of COVID-19 Symptoms at 6 Months and Above
A total of 63.87% (95% CI, 53.64-74.09%) of COVID-19 patients reported at least one symptom at 6 to 12 months, which dropped to 58.89% (95% CI, 45.87-71.91%) at 12 months and above. COVID-19 patients are at risk for long-term symptoms from multiple systems, as shown in Table 1 and Figure 2. The included studies were observational cohort studies describing the long-term consequences of COVID-19 at follow-up 6 months and above, which involved 10,945 cases of SARS-CoV-2 infection. A total of 26 studies described COVID-19 consequences at 6-12 months' follow-up and 19 studies described COVID-19 consequences at 12 months and above. The majority of the included studies were of great methodological rigor (i.e., 7-9 stars on the Newcastle-Ottawa Scale); only 2 included studies had 6 stars, mainly due to the insufficient comparability between the exposed cohort and unexposed cohort. The characteristics of the included studies are shown in Supplementary Table S2.

Pooled Prevalence of COVID-19 Symptoms at 6 Months and Above
A total of 63.87% (95% CI, 53.64-74.09%) of COVID-19 patients reported at least one symptom at 6 to 12 months, which dropped to 58.89% (95% CI, 45.87-71.91%) at 12 months and above. COVID-19 patients are at risk for long-term symptoms from multiple systems, as shown in Table 1 and Figure 2.

Pooled Prevalence of Pulmonary Functional Test Results after COVID-19 at 6 Month Above
Lung function tests showed that some participants had varying degrees of red in lung function after COVID-19 at 6 months and above. For example, during the up at 6 to 12 months, FEV1/FEV < 70% occurred in 22.86% of participants (95% C 36.77%). Abnormal pulmonary diffuse function (DLCO < 80%) was noteworthy occurred in 49.1% of the participants (95% CI, 33.27-64.9%) at 6-12 months' follow-

The Impact of COVID-19 on Quality of Life
Assessed by the EQ-5D-5L test, the quality of life of people with COVID-19 was affected in the long term, as shown in Table 2. Pain or discomfort and anxiety and depression were the most common, and personal care problems were the least common. At 6 to 12 months' follow-up, 33

Quality Evaluation, Risk of Bias, and Publication Bias
We evaluated the quality of all 40 included studies according to the Newcastle-Ottawa quality assessment scale, 38 of them were of good quality and had a low risk of bias (≥7 stars), and 2 were of moderate quality and moderate risk of bias (6 stars), as shown in Supplementary Table S3. We evaluated the publication bias of the included studies based on Harbord's modified test, and the p values of Harbord's modified test for all the meta-analyses were higher than 0.1, indicating that there was no publication bias.

GRADE Evidence Evaluation
We evaluated the evidence quality of all long-term health consequences of COVID-19 using the GRADE approach. The 40 included studies were all observational studies. After a detailed evaluation of 75 long-term COVID-19 consequences at 6-12 months' followup, a total of 3 outcomes were identified as high-quality evidence, 19 outcomes were identified as moderate-quality evidence, and the remaining 53 outcomes were identified as low-quality evidence. After the assessment of 57 long-term COVID-19 consequences at 12 months' follow-up and above, a total of 4 outcomes were identified as high-quality evidence, 17 outcomes were identified as moderate-quality evidence, and the remaining 36 outcomes were identified as low-quality evidence. The detailed results are shown in Supplementary Tables S4 and S5.

Discussion
Nowadays, COVID-19 continues to ravage the world, and although the infection of the pandemic Omicron variant may be mild [29,59], that does not mean we should relax our guard. Currently, the data on the effects of COVID-19 are growing rapidly. These data suggested that even if COVID-19 patients fully recover, they may face the risk of a variety of mid-and long-term effects [60]. Our systematic review and meta-analysis of 40 cohort studies involving 10,945 cases of SARS-CoV-2 infection provide the pooled prevalence (PP) of long-term consequences of COVID-19 at 6 months and above, and we compared subgroups stratified by follow-up period, severity of COVID-19, and gender. Understanding the long-term sequelae of COVID-19 is key to early intervention, treatment, and vaccination deployment. Previous studies have looked at the COVID -19 consequences at three months or longer [61]. Our study included a longer follow-up period of 6 months or more and a more comprehensive scope, including general, cardiovascular, respiratory, gastrointestinal, and psychiatric system symptoms, as well as the evaluation of medical imaging, lung function, and quality of life.
Consistent with previous studies, the proportion of patients with at least one symptom was as high as 60% at 6 months' follow-up, and showed a decreasing trend over time [59]. However, it should not be ignored that the proportion of patients with at least one symptom was still more than 50% when followed up at 12 months or more. In Lombardo's study, the proportion was higher, at more than 80%, but other studies have reported a lower proportion (about 40%) [29,59]. This suggests that COVID-19 may lead to sustained effects on organs, and the inconsistent results of 12-month follow-up studies suggest that more original studies on the long-term sequelae of COVID-19 are needed.
Available data analyses have shown that respiratory symptoms were common in long COVID-19, and a high PP of persistent dyspnea is of concern. A French study found that hyperventilation syndrome was common in COVID-19 patients (34%) [60], which may be related to the occurrence of persistent dyspnea. People with COVID-19 could suffer from varying degrees of respiratory damage. The available data showed that mild dyspnea was one of the most common symptoms in long-term COVID, and the proportions of CT abnormity and abnormal pulmonary diffuse function were reduced over time, which indicates that lung damage could be improved. In addition, we should also consider the impact of underlying respiratory conditions. In one meta-analysis, COPD patients with COVID-19 had a greater risk of severe disease than the non-COPD group (calculated Risk Ratio, RR = 1.88, 95% CI, 1.4-2.4) [61]. Another study found that COPD was associated with persistent symptoms at 12 months and above (OR = 10.74, p < 0.05) [59]. For people with such underlying diseases, COVID-19 sequelae may increase their burden.
The PPs of diffuse lung function impairment (DLCO < 80%) and pulmonary fibrosis were higher at long-term follow-up, but it is encouraging that this lung damage caused by COVID-19 did not appear to develop over time. In this study, diffuse lung function impairment decreased from 50% at 6-12 months to 30% at 12 months at least, and pulmonary fibrosis decreased from 66% to 14%. A study of COVID-19 patients discharged for 12 months showed no further development of pulmonary fibrosis and progressive pulmonary interstitial changes during long-term follow-up [42]. However, it should be cautioned that the repair of pulmonary fibrosis injury may bring a great burden to patients [62].
Health-related quality of life (HRQoL) is an important indicator to evaluate the impact of diseases on patients' physical, psychological, and social fields [63], and the EQ-5D-5L questionnaire is one of the most commonly used tools [64]. Our results suggest that COVID-19 patients may have long-term problems with quality of life and mental well-being, and that women are more likely to be affected than men. This could be because women, more than men, tend to take care of the family and the housework, and the job and income loss have caused women to face an economic crisis at the same time, as well as facing a larger burden of unpaid care [65]. In addition, women's exposure to domestic violence has increased because of social restrictions and isolation [66].
Since the long-term effects of COVID-19 are still unclear, the best way to reduce the consequences is to avoid infection, for which vaccination is important. In addition, improving COVID-19 screening and diagnosis capabilities can help the detection and treatment as early as possible. We should pay more attention to women's mental health and give them more psychological support, even interventions when necessary, since they are more likely to have psychological problems compared to men. In addition to the original research on the long-term effects of COVID-19, articles on the effects of vaccines on the consequences of COVID-19 are also needed.
There is currently a lack of RCTs to evaluate interventions for the long-term impact of COVID-19. This study focused on the meta-analysis of the clinical features of the long-term impacts of COVID-19. Research studies on intervention for long-term effects of COVID-19 are recommended in the future to provide evidence-based medical evidence of high GRADE quality for the development of clinical guidelines.
Our study has some limitations. First, due to limited data, some COVID-19 consequences could only analyze the PP at either 6-12 months' follow-up or at 12 months and more, not both. In addition, the heterogeneity of the PP for long-term COVID-19 effects was high, which may be related to age and gender differences.

Conclusions
Our results show that 63.87% of COVID-19 patients had at least one type of COVID-19 consequences at 6-12 months' follow-up after recovery or discharge, and 58.89% of patients continued to suffer at 12 months' follow-up and above. The most common symptoms were fatigue or muscle weakness (6-12 m: PP = 54.21%, ≥ 12 m: PP = 34.22%) and mild dyspnea (mMRC = 0) (6-12 m: PP = 74.60%, ≥12 m: PP = 80.64%). Anxiety and depression (6-12 m: PP = 33.49%, ≥12 m: PP = 35.40%) and pain or discomfort (6-12 m: PP = 33.26%, ≥12 m: PP = 35.31%) became the two most common problems affecting patients' quality of life. Our findings suggest significant long-term impacts of COVID-19 on health and quality of life, and as waves of ASRS-CoV-2 infections emerge, the long-term effects of COVID-19 will not only increase the difficulty of the care for COVID-19 survivors and setting public health policy but also might lead to another public health crisis following the current pandemic, which would also increase the global long-term burden of disease. Therefore, the long-term effects of COVID-19 should not be ignored, and it is crucial to provide a more comprehensive and scientific basis for COVID-19 survivors to guide long-term care, rehabilitation, surveillance, and prevention measures, and to set public health policy for healthcare facilities.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph19116865/s1, Table S1: Template of primary data extraction; Table S2: Characteristic of the included studies; Table S3: Risk of bias and quality of included studies assessed by Newcastle-Ottawa quality assessment Scale (NOS); Table S4: GRADE evidence evaluation results of COVID-19 consequences in 6-12 months follow-up; Table S5: GRADE evidence evaluation results of COVID-19 consequences in 12 months and above follow-up.  The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the paper. No payment was received by any of the co-authors for the preparation of this article.