Unexpected Hepatitis B Virus Infection After Liver Transplantation — United States, 2014–2019

Unexpected donor-derived hepatitis B virus (HBV) infection is defined as a new HBV infection in a recipient of a transplanted organ from a donor who tested negative for total antihepatitis B core antibody (total anti-HBc), hepatitis B surface antigen (HBsAg), and HBV DNA* before organ procurement. Such infections are rare and are associated with injection drug use among deceased donors (1). During 2014-2019, CDC received 20 reports of HBV infection among recipients of livers from donors who had no evidence of past or current HBV infection. Investigation included review of laboratory data and medical records. Fourteen of these new HBV infections were detected during 2019 alone; infections were detected a median of 38 (range = 5-116) weeks after transplantation. Of the 14 donors, 13 were hepatitis C virus (HCV)-seropositive† and had a history of injection drug use within the year preceding death, a positive toxicology result, or both. Because injection drug use is the most commonly reported risk factor for hepatitis C,§ providers caring for recipients of organs from donors who are HCV-seropositive or recently injected drugs should maintain awareness of infectious complications of injection drug use and monitor recipients accordingly (2). In addition to testing for HBV DNA at 4-6 weeks after transplantation, clinicians caring for liver transplant recipients should consider testing for HBV DNA 1 year after transplantation or at any time if signs and symptoms of viral hepatitis develop, even if previous tests were negative (2).

Unexpected donor-derived hepatitis B virus (HBV) infection is defined as a new HBV infection in a recipient of a transplanted organ from a donor who tested negative for total antihepatitis B core antibody (total anti-HBc), hepatitis B surface antigen (HBsAg), and HBV DNA* before organ procurement. Such infections are rare and are associated with injection drug use among deceased donors (1). During 2014-2019, CDC received 20 reports of HBV infection among recipients of livers from donors who had no evidence of past or current HBV infection. Investigation included review of laboratory data and medical records. Fourteen of these new HBV infections were detected during 2019 alone; infections were detected a median of 38 (range = 5-116) weeks after transplantation. Of the 14 donors, 13 were hepatitis C virus (HCV)-seropositive † and had a history of injection drug use within the year preceding death, a positive toxicology result, or both. Because injection drug use is the most commonly reported risk factor for hepatitis C, § providers caring for recipients of organs from donors who are HCV-seropositive or recently injected drugs should maintain awareness of infectious complications of injection drug use and monitor recipients accordingly (2). In addition to testing for HBV DNA at 4-6 weeks after transplantation, clinicians caring for liver transplant recipients should consider testing for HBV DNA 1 year after transplantation or at any time if signs and symptoms of viral hepatitis develop, even if previous tests were negative (2). * HBsAg and HBV DNA are laboratory evidence of current infection with HBV.
Total anti-HBc indicates past or current infection with HBV. Recipients of a liver from a donor with isolated total anti-HBc positive results can develop reactivation of hepatitis B after transplantation. † Donors who are HCV-seropositive include both HCV-viremic (anti-HCV-seropositive and HCV RNA-positve) and HCV-nonviremic (anti-HCV-seropositive and HCV RNA-negative) donors as described in https://pubmed.ncbi.nlm.nih.gov/28556422/. § https://www.cdc.gov/hepatitis/statistics/SurveillanceRpts.htm All suspected unexpected cases of donor-derived hepatitis B in the United States are reported to the Organ Procurement and Transplantation Network for review by the Ad Hoc Disease Transmission Advisory Committee. Suspected cases are referred to CDC to investigate whether donor-derived disease transmission occurred and identify interventions to prevent transmission and improve outcomes (1,2). Confirmed cases were defined as unexpected, new, ¶ reproducible laboratory evidence of HBV infection (HBsAg or HBV DNA) occurring in liver recipients after transplantation that were reported to CDC during 2014-2019. All recipients who received organs from the same donor as the liver recipient were evaluated for donor-derived HBV infection using the same criteria. Available ¶ New infection with HBV is defined as a positive viral detection test (HBsAg or HBV DNA) in an organ recipient without evidence for HBV infection (anti-HBc, HBsAg, or HBV DNA) preceding transplantation.
archived donor serum, plasma, or liver biopsy samples were tested for HBV DNA. State and local health departments shared information about recipient behavioral risk factors and outbreaks of health-care-associated HBV infection. During 2014-2019, CDC investigated 30 suspected cases of unexpected, donor-derived HBV infection among liver recipients. Ten suspected cases were excluded because the recipients had nonreproducible HBV DNA (six), or false-positive total anti-HBc (two) or HBsAg (two) results. Twenty confirmed cases were included.
Median age at death of the 20 donors was 31 years (range = 20-46 years); 11 were male, and 19 were White. The most common cause of death was drug intoxication. Injection drug use and positive toxicology were each reported for 18 donors (Table). Sixteen donors, including 13 of 14 reported in 2019, were HCV antibody (anti-HCV)-seropositive; among these 13 donors, 12 had positive drug toxicology, 12 had a history of injection drug use, and 11 had both. Stimulants (cocaine or amphetamines) were the most common substances identified by toxicology screening. HBV DNA was detected in one archived donor serum sample and one archived liver biopsy specimen.
New HBV infection was identified in 18 liver and two liverkidney recipients at a median of 41 weeks after transplantation (range = 5-116 weeks). Among cases reported during 2019, hepatitis B test conversion was first identified at a median of 38 weeks after transplantation ( Figure). None of 31 recipients of nonliver organs** from the 20 donors developed a new infection with hepatitis B. No behavioral risk factors or health care-associated hepatitis B outbreaks were reported in association with any case. Hepatitis B vaccination status was unavailable for the majority of recipients.

Discussion
HBV infection among transplant recipients can occur from reactivation of previous HBV infection (3), primary infection after transplantation, or donor-derived transmission (1). This report provides evidence that transmission of HBV from donors occurred despite negative organ donor HBV DNA, HBsAg, and total anti-HBc results before organ procurement. Among 14 cases reported during 2019, all donors but one were HCV-seropositive with a history of injection drug use, a positive toxicology result, or both. Clinicians caring for liver recipients, particularly those from donors with positive anti-HCV serology or a history of injection drug use, should maintain awareness of delayed HBV presentation and consider testing for HBV DNA at 1 year after transplantation or at any time if signs and symptoms of viral hepatitis develop, even if prior tests were negative (2). ** Analysis included nonliver organ recipients without evidence of hepatitis B infection (total anti-HBc, HBsAg, or HBV DNA) before transplantation. Twenty single kidney, six heart, and four bilateral lung recpients and one kidney-pancreas recipient received negative test results for HBsAg or HBV DNA after transplantation at the time of the investigation. Seven nonliver organ recipients with previous evidence of hepatitis B infection were excluded from this analysis, including six single kidney recipients and one bilateral lung recipient. Donors might have been exposed to HBV through injection drug use shortly before death; thus, organ procurement might have occurred during the eclipse period, † † before HBV DNA was detectable in donor serum. During the eclipse period, † † Eclipse period is defined as the 1-12 weeks between exposure to HBV and first detection of HBV DNA in serum.

TABLE. Demographic and clinical characteristics and risk behaviors of deceased organ donors* reported to CDC because of hepatitis B virus infection in liver transplant recipients after transplantation -United States, 2014-2019
HBV enters the hepatocyte nucleus and forms covalently closed circular DNA, which endures throughout the life of the nondividing hepatocyte (4). Therefore, liver recipients should be more likely than nonliver organ recipients to experience HBV infection from donors with eclipse period infection. An alternative hypothesis is that HCV coinfection suppressed

Summary
What is already known about this topic?
Unexpected donor-derived hepatitis B virus (HBV) infection after organ transplantation is rare and is associated most commonly with donor injection drug use.
What is added by this report?
During 2019, the Organ Procurement and Transplantation Network and CDC received an increased number of reports of HBV infection among liver recipients from HBV-negative donors; 12 of 14 implicated donors had evidence of recent injection drug use, and 13 donors were hepatitis C virus (HCV)-seropositive.
What are the implications for public health practice?
Providers caring for recipients of organs from donors who are HCV-seropositive or who recently injected drugs should maintain awareness of infectious complications of drug use and monitor recipients accordingly.
HBV replication in certain donors, resulting in occult HBV infection. In 20% of HBV/HCV coinfections, patients can test negative for all HBV serum markers (5). Subsequent immunosuppression or treatment for HCV infection among liver recipients might lead to reactivation of HBV infection (5) after transplantation. The observed interval (median = 41 weeks) between transplantation and diagnosis of HBV infection in these cases is similar to the prolonged interval between transplantation and reactivation of hepatitis B infection among recipients of a liver from a donor who was total anti-HBc seropositive (3).
In the United States, liver transplants from HCV-seropositive donors increased from 308 in 2014 to 644 in 2018, and liver transplants from HCV RNA-positive donors increased from 236 in 2015 to 418 in 2018 (6). The national rate of drug overdose deaths per 100,000 population § § increased during 2012-2018 from 1.4 to 4.5 for cocaine, and from 0.8 to 3.9 for psychostimulants, including amphetamines (7). Deaths related to synthetic opioids also increased during that time frame (7). ¶ ¶ Injection of cocaine (8) or methamphetamine (9) and highrisk sexual behavior (8) have been reported in association with hepatitis B outbreaks. These data indicate that the increased number of unexpected donor-derived HBV infections among liver recipients during 2019 might be related to changes in patterns of stimulant use and associated behaviors, or to increased transplantation of organs from anti-HCV-seropositive donors who injected drugs. The most common risk factor for hepatitis B and hepatitis C is injection drug use. § § Adjusted to the 2000 U.S. standard population. ¶ ¶ The trend toward increasing deaths from stimulants and opioids continued into 2019. https://www.cdc.gov/mmwr/volumes/70/wr/mm7006a4. htm?s_cid=mm7006a4_w The findings in this report are subject to at least four limitations. First, detection of infection after transplantation is dependent on testing and reporting by transplant centers. The 2013 Public Health Service guidelines (10) recommended risk-based recipient screening for hepatitis B after transplantation. However, the timing and frequency of recipient testing after transplantation might have varied during the timeframe of this study by year, transplant center, organ type, or the donor's hepatitis C status. The impact on these findings cannot be quantified but might result in underestimation of donorderived HBV infections. Second, previous recommendations (10) did not specify how hepatitis B testing of recipients should be accomplished before transplantation. Because of incomplete test results before transplantation, the presence of resolved or occult HBV infection before transplantation cannot be ruled out for certain recipients. Third, archived liver biopsy specimens were unavailable for the majority of donors. If stored correctly, liver tissue is the most likely specimen to have detectable HBV DNA during the eclipse period, which might confirm donor-derived infection. Finally, despite efforts to ascertain risk factors, risk behaviors for organ recipients might have been underreported, resulting in overestimation of donor-derived infections.
Early detection of donor-derived HBV infection is important for preventing hepatitis B-related complications among organ recipients and unintended transmission to their contacts. Recipients should be offered hepatitis B vaccination and hepatitis B testing (including total anti-HBc, HBsAg, and HBV surface antibody) before transplantation and HBV DNA testing at 4-6 weeks after transplantation (2). Additional testing for HBV DNA 1 year after transplantation (2) should be considered for liver transplant recipients, especially if the donor had risk factors for hepatitis B, including injection drug use or positive HCV serology. Recipients with signs or symptoms of liver injury after transplantation should be tested for viral hepatitis, even if previous hepatitis B or hepatitis C testing was negative (2). More broadly, providers caring for recipients of organs from donors who recently injected drugs or are HCV-seropositive should maintain awareness of infectious complications of drug use and monitor recipients accordingly. As of June 30, 2021, 33.5 million persons in the United States had received a diagnosis of . Although most patients infected with SARS-CoV-2, the virus that causes COVID-19, recover within a few weeks, some experience post-COVID-19 conditions. These range from new or returning to ongoing health problems that can continue beyond 4 weeks. Persons who were asymptomatic at the time of infection can also experience post-COVID-19 conditions. Data on post-COVID-19 conditions are emerging and information on rehabilitation needs among persons recovering from COVID-19 is limited. Using data acquired during January 2020-March 2021 from Select Medical* outpatient rehabilitation clinics, CDC compared patient-reported measures of health, physical endurance, and health care use between patients who had recovered from COVID-19 (post-COVID-19 patients) and patients needing rehabilitation because of a current or previous diagnosis of a neoplasm (cancer) who had not experienced COVID-19 (control patients). All patients had been referred to outpatient rehabilitation. Compared with control patients, post-COVID-19 patients had higher age-and sex-adjusted odds of reporting worse physical health (adjusted odds ratio [aOR] = 1.8), pain (aOR = 2.3), and difficulty with physical activities (aOR = 1.6). Post-COVID-19 patients also had worse physical endurance, measured by the 6-minute walk test † (6MWT) (p<0.001) compared with control patients. Among patients referred to outpatient rehabilitation, those recovering from COVID-19 had poorer physical health and functional status than those who had cancer, or were recovering from cancer but not COVID-19. Patients recovering from COVID-19 might need additional clinical support, including tailored physical and mental health rehabilitation services.

Outcomes Among Patients Referred to Outpatient Rehabilitation Clinics
Data were obtained from electronic health records (EHRs) of patients referred to Select Medical's outpatient rehabilitation clinics during January 2020-March 2021. Epidemiologic, clinical, and functional data from 1,295 post-COVID-19 patients and 2,395 control patients were examined. Post-COVID-19 patients were defined as those who were referred to a Select Medical facility for post-COVID-19 physical rehabilitation. * Data used were from Select Medical, a network of rehabilitation clinics in 36 states and the District of Columbia. https://www.selectmedical.com/ † https://www.thoracic.org/statements/resources/pfet/sixminute.pdf Control patients, defined as those needing rehabilitation for a current or previous diagnosis of cancer with no history of an International Classification of Diseases, Tenth Revision (ICD-10) COVID-19 diagnosis code, § were referred to a Select Medical cancer rehabilitation program. This control population was chosen because patients in this group completed the same initial evaluations as patients referred for post-COVID-19 rehabilitation. Information on type of cancer or interval since diagnosis was not available. Patient data were collected from EHRs and initial clinical evaluation, which included selfreported health measures and a 6MWT. At intake, self-reported measures and clinical evaluations were administered for health, physical endurance, and health care use.
Using validated scales, CDC assessed patients' mental and physical health, functional health, social participation ability, applied cognition, and physical endurance with Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health (version 1.2; National Institutes of Health), PROMIS Physical Function, PROMIS Ability, ¶ Quality of Life in Neurologic Disorders (Neuro-QoL),** and the 6MWT, † † respectively. For self-reported item-level data, five-point Likert scales were recoded to proportions. T-scores § International Classification of Diseases, Tenth Revision codes used to examine potential post-COVID condition were J96.01, M62.81, R.26.2, R26.89 R53, R53.1, and R53.83. ¶ PROMIS items use a Likert-type response scale (https://commonfund.nih.gov/ promis/index). The 10 PROMIS items used in this analysis included overall self-rated health; overall quality of life; overall physical health; overall mental health; and individual items on fatigue, pain, emotional distress, and social activities and roles.
Most questions asked about a person's experience "in general," with items on fatigue, pain, and emotional problems experienced during the past 7 days. Psychometric evaluation of the PROMIS global health items were based on two global physical health (GPH) and global mental health (GMH) scales. The PROMIS GPH scale included four items that rated overall physical health (physical functioning, physical activities, pain, and fatigue). GPH and GMH total raw scores were computed by summing item scores that ranged from 1 to 5, such that higher scores reflected better functioning and are then rescaled to a mean of 50 and an SD of 10 using nationally normative data from the U.S. general population. The estimated correlation between the GPH and GMH was 0.63. ** Neuro-QoL is a set of self-report measures that assesses the health-related quality of life of adults with neurologic disorders. Neuro-QoL AC-GC assesses perceived difficulties in everyday cognitive abilities, such as memory, attention, and decisionmaking. https://www.healthmeasures.net/explore-measurement-systems/neuro-qol † † Physical endurance was assessed using the 6-minute walk test. A poor 6-minute walk distance (e.g., <300 m) might have prognostic value (i.e., usually associated with an increased risk of mortality), and a change of 14.0 to 30.5 m might be clinically relevant.
were computed for composite measures of physical and mental health, social participation ability, and applied cognition, where the summed raw scores were converted to T-scores based on standardized scoring tables; T-scores were designed to have a mean of 50 and a standard deviation (SD) of 10 for the general adult population Logistic regression analysis, adjusted for age and sex, was used to examine differences in patient-reported measures of health, physical endurance, and health care use between post-COVID-19 and control patients. § § All analyses were conducted using SAS (version 9.4; SAS Institute). This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy. ¶ ¶ Post-COVID-19 patients referred for rehabilitation services differed from control patients by several characteristics, including sex, age, race, ethnicity, employment status, health insurance coverage, and U.S. Census region ( Table 1). Compared with control patients, post-COVID-19 patients were more likely to be male, younger, in the labor force, insured by a commercial plan or a worker's compensation plan, and less likely to be covered by Medicaid or Medicare (Table 1). Post-COVID-19 patients were more likely to have received a diagnosis of generalized muscle weakness or fatigue (72.7% versus 42.3%) and patient-reported symptoms of generalized muscle weakness, malaise, and fatigue (69.0% versus 59.7%) ( Table 2).
Compared with control patients, post-COVID-19 patients had higher prevalences of reported fair or poor general health (32.9% versus 25.4%), poorer physical health (44.1% versus 32.6%), pain level ≥7 (on a scale of 0-10) (40.4% versus 24.8%), and difficulty with physical activities (32.3% versus 24.2%) ( Table 3). Post-COVID-19 patients also reported a higher prevalence of fair or poor overall mental health than control patients (19.1% versus 15.3%). Post-COVID-19 patients and control patients reported more challenges with applied cognition as indicated by T-scores (42.2 versus 41.2), both approximately one SD below the normative sample with which the scale was developed. Post-COVID-19 patients also demonstrated reduced physical endurance on the 6MWT compared with control patients (distance of 303 m versus 377 m; p<0.001) and reported increased difficulty completing chores (38.2% versus 25.2%), navigating stairs (40.2% versus 18.3%), running errands or shopping (34.3% versus 16.0%), and walking for 15 minutes (38.2% versus 16.6%). Compared with control patients, post-COVID-19 patients also reported more difficulty doing usual work or work at home (37.2% versus 20.4%) and challenges in ability to participate in activities § § Other demographic variables besides sex and age had substantial proportions of missing data (26%-75%); therefore, these variables were not included in the analysis.

Summary
What is already known about this topic?
COVID-19 patients might experience symptoms that persist months after initial infection.
What is added by this report?
Compared with control patients enrolled in a cancer rehabilitation program, adult post-COVID-19 patients referred for rehabilitation services reported poorer physical health and being less able to engage in physical activities and activities of daily living. Patients recovering from COVID-19 also had significantly higher health care use than control patients.
What are the implications for public health practice?
Patients recovering from COVID-19 might require tailored physical and mental health rehabilitation services.

Discussion
Among patients referred to Select Medical's outpatient rehabilitation clinics during January 2020-March 2021 (during the COVID-19 pandemic), patients who previously had COVID-19 reported poorer general, mental, and physical health (i.e., overall physical health, physical activities, and pain), and functioning (i.e., physical and social, such as ability to do chores, usual work, or activities with friends) compared with patients with no previous diagnosis of COVID-19 referred for cancer rehabilitation. Also, post-COVID-19 patients did not perform as well as control patients on a measured assessment of physical functioning (6MWT). Finally, post-COVID-19 patients used more rehabilitative services than control patients. These findings indicate that among patients referred to outpatient rehabilitation, those recovering from COVID-19 might have poorer physical health and functional status than do patients with cancer but not COVID-19 and could benefit from additional clinical support, including tailored physical and mental health rehabilitation services.
The identification of poorer physical health among post-COVID-19 patients is consistent with a previous study that found that 92% of post-COVID-19 patients had diagnoses potentially related to post-COVID-19 conditions, including weakness, malaise, fatigue, respiratory failure with hypoxia, and gait abnormalities (2,3). Poorer self-reported physical and mental health is associated with long-term negative health outcomes including chronic diseases (e.g., diabetes and cardiovascular  disease), functional decline (4), and mortality (5). The lower scores on applied cognitive ability tasks suggest more subtle deficits in cognitive functioning, which might indicate the need for further evaluation and additional need for health care resources and services (6). Further, physical function, as measured by the 6MWT, has been shown to be an important outcome for assessing impact of . Additional studies have shown that patients recovering from COVID-19 have higher incidences of negative health outcomes, including poorer physical health and functional status, and might need additional clinical support such as tailored physical and mental health rehabilitation services (7,8). These findings have implications for health care systems during and after the COVID-19 pandemic (9). Postacute sequalae associated with COVID-19 have not been comprehensively described, and data from studies of long-term follow-up to provide reliable estimates of the long-term sequelae associated with COVID-19 are still emerging (6)(7)(8). Continued assessments and sex. ¶ Proportions of patients reporting "fair" or "poor" general health. ** Mental and physical health were assessed with PROMIS Scale v1.2 -Global Health (National Institutes of Health). PROMIS items all use a Likert-type response scale. Most questions ask about a person's experience "in general, " with items on fatigue, pain, and emotional problems referencing the past 7 days. The PROMIS global mental health scale includes four items that rate overall mental health (quality of life, mental health, emotional distress, and social activities and roles). The PROMIS global physical health scale includes four items that rate overall physical health (physical functioning, physical activities, pain, and fatigue). Proportions of patients reporting "fair" or "poor" health were calculated for each measure, with the exceptions of emotional problems, physical activities, pain, and fatigue. Proportions of patients reporting "often" or "always" were calculated for emotional problems; "little" or "none at all" for physical activities; and "severe" or "very severe" for fatigue. Pain was measured using a scale of 0-10 and the proportion of patients reporting ≥7 was calculated. † † Physical functional status was assessed with PROMIS Item Bank v2.0 -Physical Function-Short Form 4a. Proportions of patients reporting "with much difficulty" or "unable to do with much difficulty" were calculated for each measure. § § Social participation ability was assessed with PROMIS Item Bank v2.0 -Ability to Participate in Social Roles and Activities-Short Form 4a. Proportions of patients reporting "usually" or "always" were calculated for each measure. ¶ ¶ Applied cognition was assessed with Neuro-QOL Item Bank v1.0 -Applied Cognition -General Concerns (AC-GC)-Short Form. Neuro-QoL AC-GC assesses perceived difficulties in everyday cognitive abilities such as memory, attention, and decision-making. Proportions of patients reporting "often (once a day)" or "very often (several times a day)" were calculated for each measure. *** Total raw scores were computed by summing items scores that range from 1 to 5, such that higher scores reflect better functioning and are then rescaled to a mean of 50 and SD of 10 using nationally normative data from the U.S. general population. † † † Physical endurance was assessed using the 6-minute walk test. A poor 6-minute walk distance (e.g., <300 m) might have prognostic value (i.e., usually associated with an increased risk of mortality), and a change of 14.0 to 30.5 m might be clinically relevant.
of self-reported health data are important to characterize the sequelae of novel infectious diseases and are critical for developing cost-effectiveness estimates for lifesaving interventions, such as vaccines and other potentially important rehabilitation therapies and interventions, including physical therapy, occupational therapy, and services and therapies associated with cognitive and functional decline (9,10). The findings in this report are subject to at least six limitations. First, date of infection was not available; therefore, time-varying effects associated with infection date could not be examined. Second, data on severity of illness, including hospitalization status, were not available, precluding assessment of the impact of illness severity on post-COVID-19 conditions. Third, given the large amount of missing data (>50%) for many demographic variables (e.g., race, ethnicity, employment status, and occupation), which are common limitations in large EHR data sets, it was not possible to control for additional demographic differences. Fourth, the absence of pre-COVID-19 assessments did not permit controlling for premorbid function. Fifth, the types of cancer diagnoses and treatments were not available, which is an important consideration given heterogeneity of cancer sequelae. Similarly, assessing other comorbidities was not possible; post-COVID-19 patients might have had more underlying medical conditions (e.g., diabetes or obesity) than did control patients, which could explain poorer physical and mental health measures. However, given that patients in the post-COVID-19 group were younger and more commonly employed than were those in the control group, it is likely that these two populations are different with regard to demographic factors and the prevalence of comorbid chronic conditions. Finally, referral to physical rehabilitation depended on nonstandardized clinical judgment, which might have led to differences in patient population by group. Therefore, these results should not be interpreted to mean that post-COVID-19 patients overall had poorer physical and mental health than patients with cancer. Instead, results indicate that post-COVID-19 patients specifically referred to a large physical rehabilitation network had poorer health measures than those referred for cancer, which indicates that some patients recovering from COVID-19 had substantial rehabilitation needs.
Patients recovering from COVID-19 might experience continued poor health and could benefit from additional support and tailored physical and mental health rehabilitation services. Health care systems and providers should be prepared to recognize and meet the ongoing needs of this patient population. Efforts to increase COVID-19 vaccination could include messaging that states that preventing COVID-19 also prevents post-COVID-19 conditions with potential effects on long-term health.

Efficacy of Portable Air Cleaners and Masking for Reducing Indoor Exposure to Simulated Exhaled SARS-CoV-2 Aerosols -United States, 2021
William G. Release on the MMWR website (https://www.cdc.gov/mmwr). SARS-CoV-2, the virus that causes COVID-19, can be spread by exposure to droplets and aerosols of respiratory fluids that are released by infected persons when they cough, sing, talk, or exhale. To reduce indoor transmission of SARS-CoV-2 between persons, CDC recommends measures including physical distancing, universal masking (the use of face masks in public places by everyone who is not fully vaccinated), and increased room ventilation (1). Ventilation systems can be supplemented with portable high efficiency particulate air (HEPA) cleaners* to reduce the number of infectious particles in the air and provide enhanced protection from transmission between persons (2); two recent reports found that HEPA air cleaners in classrooms could reduce overall aerosol particle concentrations by ≥80% within 30 minutes (3,4). To investigate the effectiveness of portable HEPA air cleaners and universal masking at reducing exposure to exhaled aerosol particles, the investigation team used respiratory simulators to mimic a person with COVID-19 and other, uninfected persons in a conference room. The addition of two HEPA air cleaners that met the Environmental Protection Agency (EPA)-recommended clean air delivery rate (CADR) (5) reduced overall exposure to simulated exhaled aerosol particles by up to 65% without universal masking. Without the HEPA air cleaners, universal masking reduced the combined mean aerosol concentration by 72%. The combination of the two HEPA air cleaners and universal masking reduced overall exposure by up to 90%. The HEPA air cleaners were most effective when they were close to the aerosol source. These findings suggest that portable HEPA air cleaners can reduce exposure to SARS-CoV-2 aerosols in indoor environments, with greater reductions in exposure occurring when used in combination with universal masking.
A breathing aerosol source simulator was used to mimic a meeting participant exhaling infectious particles (source), and three breathing simulators were used to mimic a speaker and two participants exposed to these aerosol particles (receivers) (Figure 1). The methods used were similar to those used in previous studies of aerosol dispersion and transport in indoor spaces (3,4,6). The simulators were placed in a 584-ft 2 (54-m 2 ) * HEPA air cleaners consist of a filter capable of removing ≥99.97% of particles from the air and a fan or blower to draw air through the filter. HEPA air cleaners are commercially available, relatively inexpensive, and easy to use.
On July 2, 2021, this report was posted as an MMWR Early conference room with a heating, ventilation, and air conditioning ( The concentrations of 0.3 μm to 3 μm aerosol particles were measured at the mouth of each receiver using optical particle counters (Model 1.108, Grimm Technologies, Inc.) to determine the exposure of each receiver simulator to aerosol particles. When the simulators were masked, the particle counters collected aerosol samples from inside the masks (i.e., the particle counter measured the concentration of the aerosol being inhaled by the receiver simulator). For each optical particle counter, the total aerosol mass concentration was averaged over 60 minutes to determine the mean aerosol mass concentration (mean aerosol exposure) to which each receiver was exposed. Each experiment was repeated four times for a total of 20 tests. All data were analyzed using the Kruskal Wallis test to assess overall significance, followed by a Wilcoxon Rank Sum pairwise comparison with a Benjamini and Hochberg adjusted p-value for multiple comparisons. R software (version 3.6.0; R Foundation) was used to conduct all analyses. . The room ventilation system air inlets and outlets were located in the ceiling as part of the light fixtures. § The source simulator breathed continuously at 15 liters per minute, and the aerosol generator was repeatedly cycled on for 20 seconds and off for 40 seconds to avoid exceeding the range of the aerosol instruments. ¶ Two participant breathing simulators (participant receivers) had a design based on the respiratory aerosol source simulator and breathed continuously at 15 liters per minute. The speaker breathing simulator (speaker receiver) was a commercial simulator that breathed at 28 liters per minute.

Summary
What is already known about this topic? Ventilation systems can be supplemented with portable high efficiency particulate air (HEPA) cleaners to reduce the number of airborne infectious particles.
What is added by this report?
A simulated infected meeting participant who was exhaling aerosols was placed in a room with two simulated uninfected participants and a simulated uninfected speaker. Using two HEPA air cleaners close to the aerosol source reduced the aerosol exposure of the uninfected participants and speaker by up to 65%. A combination of HEPA air cleaners and universal masking reduced exposure by up to 90%.
What are the implications for public health practice?
Portable HEPA air cleaners can reduce exposure to simulated SARS-CoV-2 aerosols in indoor environments, especially when combined with universal masking.
The mean aerosol concentrations for the two participant receivers and the speaker receiver were generally similar during each experiment, indicating that the air in the room was well mixed over the 60-minute test period (Table). For all assessed scenarios, use of the HEPA air cleaners significantly reduced the aerosol exposures for the two participant receivers and speaker receiver (p = 0.001) (Figure 2). Without masks, the combined mean aerosol concentrations for the two participant receivers and speaker receiver were reduced by 49% with the air cleaners in the left and right elevated positions, 52% in the left and right floor positions, 55% in the front and back floor positions, and 65% in the center floor positions. The reductions with the air cleaners in the center floor position were higher than those with the air cleaners in the left/right or front/back positions (p<0.01). The aerosol concentrations when the air cleaners were in the left and right floor, left and right elevated, and front and back floor position results did not differ significantly from one another. Without the HEPA air cleaners, universal masking reduced the combined mean aerosol concentration by 72% (p<0.001). When both universal masking and the HEPA air cleaners were used, the combined mean concentrations for the two participant receivers and the speaker decreased by as much as 90% (p<0.001) (Table).

Discussion
In this study, the use of HEPA air cleaners in a conference room significantly reduced the exposure of nearby participants and a speaker to airborne particles produced by a simulated infected participant. The air cleaners were most effective when they were located in the center of the room close to the aerosol source. Moreover, the combination of HEPA air cleaners and universal masking was more effective than was either intervention alone. The use of masks without air cleaners reduced the aerosol exposure of the receivers by 72%, and the use of air cleaners without masks reduced the exposure by up to 65%. When used together, the HEPA air cleaners and masks reduced exposure to respiratory aerosols by up to 90%. These findings suggest that the use of portable HEPA air cleaners and universal masking can each reduce exposure to simulated SARS-CoV-2 aerosols in indoor environments, with larger reductions occurring when air cleaners and masking are used together.
Ventilation is a well-established method for reducing potential exposures to infectious aerosols (7). By removing airborne particles from a room, ventilation systems can reduce exposures that occur by inhalation of infectious aerosols, deposition on susceptible mucous membranes, or conveyance to mucous membranes by contaminated hands. However, in most nonclinical settings, ventilation systems are designed only with sufficient airflow to provide fresh air while maintaining comfortable temperature and humidity levels; these systems  typically are not designed to have the much higher airflow rates that are needed to reduce disease transmission (8). During the ongoing pandemic, public health and professional organizations have provided guidance for increasing ventilation and air filtration to decrease the spread of SARS-CoV-2 (2,9,10). One recommended option, especially when existing HVAC systems might be insufficient, is adding portable HEPA air cleaners to rooms (2). The results of this study support the use of portable HEPA air cleaners to reduce exposure to airborne particles.
The findings in this report are subject to at least five limitations. First, the dispersion of aerosols in a room depends upon air currents, which are unique to each setting. In this study, the conference room air was well mixed, which helped transport aerosols to the air cleaners. In rooms with poor air mixing and potential stagnation zones, air cleaners might be less effective. Airflow patterns in real-world settings such as classrooms will vary among buildings and rooms, and rooms of different dimensions and with different ventilation rates will also have different airflow patterns. Second, the aerosol source manikin in this study was kept in one fixed location. In reality, potentially infectious occupants could be anywhere in the room and might move around the room occasionally. Third, this study only used one source manikin and three receiver manikins; additional sources and receivers could change the dynamics of aerosol dispersion within a room. Fourth, the study was limited to aerosol particles of 0.3 μm to 3 μm in size, which are small enough to remain airborne for an extended time but large enough to carry pathogens. However, particles outside this size range would behave differently. Finally, the study only assessed aerosol exposure; it did not directly examine disease transmission. Although the study provides useful information about the dynamics of respiratory aerosol particles and the effects of HEPA air cleaners and universal masking, many other factors are also important for disease transmission, including the amount of virus in the particles, how long the virus survives in air, and the vaccination status of the room occupants.
Portable HEPA air cleaners offer a simple means to increase the filtration of aerosol particles from a room without modifying the existing building ventilation system (2). The optimal location for HEPA air cleaners will depend upon the unique conditions in each room, but they are likely to be most effective when they are placed as close to the occupants as is practicable. Larger reductions in exposure occur when air cleaners are used in combination with universal masking. These On July 6, 2021 this report was posted as an MMWR Early Release on the MMWR website (https://www.cdc.gov/mmwr).
In December 2020, the Food and Drug Administration (FDA) issued Emergency Use Authorizations (EUAs) for the Pfizer-BioNTech COVID-19 (BNT162b2) vaccine and the Moderna COVID-19 (mRNA-1273) vaccine, † and the Advisory Committee on Immunization Practices (ACIP) issued interim recommendations for their use in persons aged ≥16 years and ≥18 years, respectively. § In May 2021, FDA expanded the EUA for the Pfizer-BioNTech COVID-19 vaccine to include adolescents aged 12-15 years; ACIP recommends that all persons aged ≥12 years receive a COVID-19 vaccine. Both Pfizer-BioNTech and Moderna vaccines are mRNA vaccines encoding the stabilized prefusion spike glycoprotein of SARS-CoV-2, the virus that causes COVID-19. Both mRNA vaccines were authorized and recommended as a 2-dose schedule, with second doses administered 21 days (Pfizer-BioNTech) or 28 days (Moderna) after the first dose. After reports of myocarditis and pericarditis in mRNA vaccine recipients, ¶ which predominantly occurred in young males after the second dose, an ACIP meeting was rapidly convened to review reported cases of myocarditis and pericarditis and discuss the benefits and risks of mRNA COVID-19 vaccination in the United States. Myocarditis is an inflammation of the heart muscle; if it is accompanied by pericarditis, an inflammation of the thin tissue surrounding the heart (the pericardium), it is referred to as myopericarditis. Hereafter, myocarditis is used to refer to myocarditis, pericarditis, or myopericarditis. On June 23, 2021, after reviewing available evidence including that for risks of myocarditis, ACIP determined that the benefits of using mRNA COVID-19 vaccines under the FDA's EUA clearly outweigh the risks in all populations, including adolescents and young adults. The EUA has * These authors contributed equally to this work. † All EUA documents for COVID-19 vaccines, including fact sheets, are available at https://www.fda.gov/emergency-preparedness-and-response/coronavirusdisease-2019-covid-19/covid-19-vaccines. § ACIP recommendations for all COVID-19 vaccines are available at https:// www.cdc.gov/vaccines/hcp/acip-recs/vacc-specific/covid-19.html. ¶ COVID-19 Vaccine Safety Technical Work Group Reports are available at https://www.cdc.gov/vaccines/acip/work-groups-vast/index.html. been modified to include information on myocarditis after receipt of mRNA COVID-19 vaccines. The EUA fact sheets should be provided before vaccination; in addition, CDC has developed patient and provider education materials about the possibility of myocarditis and symptoms of concern, to ensure prompt recognition and management of myocarditis.
Since June 2020, ACIP has convened 15 public meetings to review data on COVID-19 epidemiology and use of COVID-19 vaccines. The ACIP COVID-19 Vaccines Work Group, comprising experts in infectious diseases, vaccinology, vaccine safety, public health, and ethics, has held weekly meetings since April 2020 to review COVID-19 surveillance data, evidence for vaccine efficacy and safety, and implementation considerations for COVID-19 vaccination programs. After reports of myocarditis, the work group met twice to review clinical trial and postauthorization safety data for myocarditis after receipt of mRNA COVID-19 vaccines. The work group also reviewed a benefit-risk assessment of myocarditis events after receipt of mRNA COVID-19 vaccines, considering recent epidemiology of COVID-19 and sequelae of COVID-19, including myocarditis and multisystem inflammatory syndrome in children (MIS-C).** The ACIP COVID-19 Vaccines Safety Technical (VaST) Work Group, comprising independent vaccine safety expert consultants, had also reviewed safety data on myocarditis after receipt of mRNA COVID-19 vaccines at its weekly meetings. The findings from the VaST and the ACIP COVID-19 Vaccines Work Group assessments, including a summary of the data reviewed, were presented to ACIP during its meeting on June 23, 2021.
Myocarditis typically occurs more commonly in males than in females, and incidence is highest among infants, adolescents, and young adults (1,2). The clinical presentation and severity of myocarditis vary among patients. Symptoms typically include chest pain, dyspnea, or palpitations, although other symptoms might be present, especially in younger children (3). Diagnostic evaluation might reveal an elevated troponin level or abnormal findings on electrocardiogram, echocardiogram, or cardiac magnetic resonance imaging (Table 1). Supportive therapy is ** https://www.cdc.gov/mis/hcp/index.html a mainstay of treatment, with targeted cardiac medications or interventions as needed. Current guidelines from the American Heart Association and American College of Cardiology recommend exercise restriction until the heart recovers. † † As of June 11, 2021, approximately 296 million doses of mRNA COVID-19 vaccines had been administered in the United States, with 52 million administered to persons aged 12-29 years; of these, 30 million were first and 22 million were second doses. Within the Vaccine Adverse Event Reporting System (VAERS) (4), the national vaccine safety passive monitoring system, 1,226 reports of myocarditis after mRNA vaccination were received during December 29, 2020-June 11, 2021. Among persons with reported myocarditis after mRNA vaccination, the median age was 26 years (range = 12-94 years), with median symptom onset interval of 3 days after vaccination (range = 0-179). Among 1,194 reports for which patient age was known, 687 were among persons aged <30 years and 507 were among persons aged ≥30 years; of 1,212 with sex reported, 923 were male, and 289 were female. § § Among 1,094 patients with number of vaccine doses received reported, 76% occurred after receipt of dose 2 of mRNA vaccine; cases were reported after both Pfizer-BioNTech and Moderna vaccines. Informed by early reports, CDC prioritized rapid review of myocarditis in persons aged <30 years reported during May 1-June 11, 2021; the 484 patient records in this subset were evaluated by physicians at CDC, and several reports were also reviewed with Clinical Immunization Safety Assessment Project investigators, ¶ ¶ including cardiologists. At the time of this report, 323 of these 484 cases were determined to meet criteria in CDC's case definitions for myocarditis, pericarditis, or myopericarditis by provider interview or medical record review ( Table 1). The median age of the 323 patients meeting CDC's case definitions was 19 years (range = 12−29 years); 291 were male, and 32 were female. The median interval from vaccination to symptom onset was 2 days (range = 0−40 days); 92% of patients experienced onset of symptoms within 7 days of vaccination. Of the 323 persons meeting CDC's case definitions, 309 (96%) were hospitalized. Acute clinical courses were generally mild; among 304 hospitalized patients with known clinical outcomes, 95% had been discharged at time of review, and none had died. Treatment data in VAERS are preliminary and incomplete; however, many patients have experienced resolution of symptoms with conservative treatment, such as receipt of nonsteroidal antiinflammatory drugs. Follow-up is † † https://www.ahajournals.org/doi/10.1161/CIR.0000000000000239?url_ ver = Z39.88-2003&rfr_id = ori:rid:crossref.org&rfr_dat = cr_pub%20%20 0pubmed#d3e785 § § Age was not reported for 32 patients, and sex was not reported for 14 patients. ¶ ¶ https://www.cdc.gov/vaccinesafety/ensuringsafety/monitoring/cisa/index.html ongoing to identify and understand longer-term outcomes after myocarditis occurring after COVID-19 vaccination. Using myocarditis cases reported to VAERS with onset within 7 days after dose 2 of an mRNA vaccine, crude reporting rates (i.e., using confirmed and unconfirmed cases) per million second dose recipients were calculated using national COVID-19 vaccine administration data as of June 11, 2021. Myocarditis reporting rates were 40.6 cases per million second doses of mRNA COVID-19 vaccines administered to males aged 12−29 years and 2.4 per million second doses administered to males aged ≥30 years; reporting rates among females in these age groups were 4.2 and 1.0 per million second doses, respectively.*** The highest reporting rates were among males aged 12−17 years and those aged 18−24 years (62.8 and 50.5 reported myocarditis cases per million second doses of mRNA COVID-19 vaccine administered, respectively). Myocarditis rates from Vaccine Safety Datalink (VSD), based on electronic health records, were also evaluated. Although numbers were too small to show rates in all subgroups by age, VSD data indicated increased risk of myocarditis in the 7 days after receipt of dose 1 or dose 2 of an mRNA COVID-19 vaccine compared with the risk 22-42 days after the second dose, particularly among younger males after dose 2 (5).
To assess the benefit-risk balance of mRNA vaccines in adolescents and young adults, ACIP reviewed an individual-level assessment that compared the benefits (i.e., COVID-19 infections and severe disease prevented) to the risks (number of cases of myocarditis) of vaccination, using methods similar to those described previously. † † † Specifically, the benefits per million second doses administered (i.e., the benefits of being fully vaccinated in accordance with the FDA EUA) were assessed, including 1) COVID-19 cases prevented based on rates the week of May 29, 2021 § § § ; 2) COVID-19 hospitalizations prevented based on rates the week of May 22, 2021 ¶ ¶ ¶ ; and 3) COVID-19 intensive care unit (ICU) admissions and deaths prevented based on the proportion of hospitalized patients who were admitted to the ICU or died.**** The risks were assessed as the number of myocarditis patients reported to VAERS that occurred within 7 days of receipt of a second dose of an mRNA COVID-19 vaccine per million second doses administered through the week of June 11, 2021. † † † † *** Data collection for race/ethnicity of myocarditis cases is ongoing. † † † https://www.cdc.gov/vaccines/covid-19/info-by-product/janssen/riskbenefit-analysis.html § § § https://covid.cdc.gov/covid-data-tracker/#demographicsovertime. Data were used for the most recent week not subject to reporting delays prior to the ACIP meeting. ¶ ¶ ¶ https://gis.cdc.gov/grasp/COVIDNet/COVID19_3.html. Data were used for the most recent week not subject to reporting delays prior to the ACIP meeting. **** https://gis.cdc.gov/grasp/COVIDNet/COVID19_5.html † † † † Because of uncertainty in the accuracy of myocarditis reporting, given that reviews are ongoing, and some cases might not have been reported yet, myocarditis reporting rates are presented as a range of values, calculated as ±10% of the observed reporting rates.
The benefit-risk assessment was stratified by age group and sex. The analysis assumed 95% vaccine effectiveness § § § § of 2 doses of a mRNA COVID-19 vaccine in preventing COVID-19 cases and hospitalization and assessed outcomes for a 120-day period. The 120-day period was selected because 1) no alternative vaccine options currently exist for persons aged <18 years or are expected to be available during this period, and 2) inputs regarding community transmission have high uncertainty beyond this period, particularly in the context of circulating variants. ¶ ¶ ¶ ¶ The benefits (prevention of COVID-19 disease and associated hospitalizations, ICU admissions, and deaths) outweighed the risks (expected myocarditis cases after vaccination) in all populations for which vaccination has been recommended. However, the balance of benefits and risks varied by age and sex because cases of myocarditis were primarily identified among males aged <30 years, and the risks of poor outcomes related to COVID-19 increase with age. Per million second doses of mRNA COVID-19 vaccine administered to males aged 12-29 years, 11,000 COVID-19 cases, 560 hospitalizations, 138 ICU admissions, and six deaths due to COVID-19 could be prevented, compared with 39-47 expected myocarditis cases after COVID-19 vaccination ( Table 2). Among males aged ≥30 years, 15,300 COVID-19 cases, 4,598 hospitalizations, 1,242 ICU admissions, and 700 deaths could be prevented, compared with three to four expected myocarditis cases after COVID-19 vaccination. This analysis did not include the potential benefit of preventing post-COVID-19 conditions, such as prolonged symptoms and MIS-C (6,7). ACIP also reviewed population-level considerations regarding vaccination. No alternatives to mRNA COVID-19 vaccines for adolescents will be available for the foreseeable future, and vaccination of adolescents offers protection against COVID-19 that can be important for returning to educational, social, and extracurricular activities. Higher levels of vaccination coverage can reduce community transmission, which can protect against development and circulation of emerging variants. Regarding health equity considerations, racial and ethnic minority groups have higher rates of COVID-19 and severe disease*****; potential changes in vaccine policy, or anything that would affect vaccination coverage for adolescents or young adults, might disproportionately affect those groups with the highest rates of poor COVID-19 outcomes.