Abstract
Background:
Direct-acting antivirals (DAAs) for treatment of hepatitis C virus (HCV) have resulted in great success through high attainment of sustained virologic response (SVR). Risk factors for DAA treatment failure are important to identify because of worsened outcomes with failure and high treatment cost.
Objective:
We sought to identify whether hospitalization during treatment affects SVR. The primary outcome was the difference in SVR at 12 weeks after treatment
Methods:
This multicenter, single health system retrospective cohort review compared achievement of SVR between patients hospitalized during DAA treatment for HCV with those not hospitalized during treatment.
Results:
Patients in the hospitalized cohort (n = 94) had more severe disease at baseline than nonhospitalized patients (n = 167) as indicated through higher Model for End-Stage Liver Disease (MELD) scores, Fibrosis-4 scores, and imaging-suggested or biopsy-confirmed cirrhosis. Patients hospitalized during treatment had lower SVR rates compared with those not hospitalized (87.2% vs 95.2%; P = 0.043) but failed to reach significance when inpatient mortality was excluded on secondary analysis (91.1% vs 95.2%; P = 0.195). Patients who were hospitalized and did not achieve SVR had higher MELD scores, were more likely to have intensive care unit stay, and had longer hospital stay compared with those who achieved SVR. Of 94 patients, 93 provided home supply of DAAs during hospitalization.
Conclusion and Relevance:
Patients hospitalized during DAA treatment for HCV had reduced rates of SVR. This reduced SVR rate may be driven by inpatient mortality and severity of liver disease. Patient education to bring home supply of medication for use during admission is an effective intervention.
Introduction
The introduction of new direct-acting antivirals (DAAs) in 2011 has radically changed the outlook of hepatitis C virus (HCV) management, and most available DAAs obtain high sustained virologic response (SVR) rates, reaching 95% in most patient groups.1-8 SVR is defined as undetectable HCV in serum at 12 weeks or longer after treatment has been completed and is considered to be cure of HCV infection. 9 Patients treated with DAAs who achieve SVR have reduced mortality, need for liver transplantation, hepatocellular carcinoma risk, liver-related hospitalizations, and health care costs.10-14 Risk factors for treatment failure have been described and include advanced liver disease, HCV genotype, hepatocellular carcinoma, and previous DAA treatment failure especially if resistance-associated substitutions are present.15-21 Failure to achieve SVR has detrimental downstream effects on both clinical outcomes and overall cost of therapy. It is, therefore, important to continue to identify additional risk factors for DAA treatment failure.
Our study aims to investigate the possible association between hospitalization during a patient’s DAA treatment course and achieving SVR. We hypothesized that patients requiring hospitalization during DAA treatment would have reduced SVR rates compared with patients who were not hospitalized. All-cause hospitalization rates for patients with chronic liver disease doubled from 2000 to 2014. 22 Patients treated with DAAs may experience hospitalization, which could in turn affect treatment response through a variety of mechanisms. Some DAAs have significant drug-drug interactions (DDIs), and during hospitalization acute care providers not familiar with DAAs may inadvertently order medications that interact with DAAs, potentially compromising efficacy. Additionally, health systems typically do not include DAAs on their inpatient formulary because these medications are expensive and are for chronic HCV disease; therefore, patients may experience interruptions in therapy during hospitalization. Because DAA treatment is often only 8 to 12 weeks in duration, detrimental drug interactions or interruptions in therapy experienced during hospitalization may affect SVR and desired clinical outcomes. Furthermore, hospitalized patients may experience changes that alter pharmacokinetic elimination and metabolism, with variations in renal and/or hepatic function. A recent study revealed that only 74% of patients with cirrhosis hospitalized for liver-related complications after DAA treatment achieved SVR 13 ; it is important to answer the question if hospitalization during DAA treatment affects or otherwise acts as a prognostic indicator of SVR failure. In addition, identifying patients at risk for hospitalization during treatment may be important to ensure that hospitalized patients have appropriate supply of their DAA regimen during admission. For these reasons, we evaluated the association and possible effects of hospitalization during DAA treatment and subsequent SVR success or failure.
Methods
This investigator-initiated, multicenter retrospective case-control study included patients who received DAA therapy for HCV in the UC Health system. Patients for both cohorts were included from January 2013 to January 2019. Patients in the investigational case cohort were those admitted for any cause to a UC Health facility, including University of Cincinnati Medical Center, West Chester Hospital, Daniel Drake Center, and Lindner Center of HOPE, during the time the patient received DAA therapy for HCV. Patients in a control cohort were those patients who received DAA therapy for HCV and were not hospitalized at a UC Health facility during the patient’s DAA treatment period. The study was approved by the University of Cincinnati Institutional Review Board.
Control patients’ charts were reviewed for confirmation of the start of DAAs, with start date as documented in the patients’ electronic medical record (EMR) through hepatology or infectious diseases clinic notes and/or institutional specialty pharmacy notes. Planned treatment duration was reviewed in pretreatment clinic notes, and completion of treatment was verified in follow-up outpatient documentation. Providers at UC Health follow the joint American Association for the Study of Liver Diseases (AASLD) and Infectious Diseases Society of America HCV treatment guidance. The confirmation of start of DAA therapy was critical because patients may have been prescribed therapy but may not have actually received it or may have encountered delay for various reasons. Control patients were confirmed to have not been admitted to a UC Health facility during DAA treatment through review of the EMR. Only hospitalizations within our health system were included. The case patients were identified through 2 reports. The first report identified any HCV DAA medication order entered in the inpatient institutional setting. Through review of patients’ EMR, they were confirmed to have been receiving DAA therapy during admission. The second report involved patients with a DAA prescription who were identified as having an admission to a UC Health facility within 365 days of the original DAA prescription. The report was necessary because patients may have been admitted but the DAAs may not have been ordered because HCV DAAs are not on hospital formulary. The 365-day postprescribing period was chosen because patients may have been prescribed a DAA but may have encountered delays in receipt of therapy. Prescriptions in Ohio expire beyond 1 year, so delays beyond 1 year would require a new prescription, which would show on our report. These patients were verified to have received DAA therapy (through hepatology or infectious diseases clinic notes and/or institutional specialty pharmacy notes) and that admission to the UC Health facility occurred during treatment. A review for duplication of case patients was completed for both reports. DAAs are nonformulary in our health system institutions, and therefore, patients are educated to provide home supply of DAAs during an admission.
Inclusion criteria were age ≥18 years old, both pretreatment and posttreatment (at least 12 weeks after treatment) HCV RNA concentrations within the EMR (unless mortality occurred during treatment), and receipt of any of the following medications or combinations (in addition to ribavirin if included as part of the regimen): sofosbuvir, simeprevir, daclatasvir, sofosbuvir/ledipasvir, sofosbuvir/velpatasvir, sofosbuvir/velpatasvir/voxilaprevir, elbasvir/grazoprevir, glecaprevir/pibrentasvir, ombitasvir/paritaprevir/ritonavir, or ombitasvir/paritaprevir/ritonavir plus dasabuvir. Initiation and completion of DAA therapy must have been verified through outpatient documentation. Length of stay must have been ≥24 hours to ensure that the patient was due to take a dose of DAA medication. Patients were excluded if pregnant.
The primary outcome was attainment of SVR as indicated by undetectable posttreatment viral load at least 12 weeks after completion. This outcome was compared between patients admitted during DAA therapy to those not admitted during DAA therapy. If a patient died during therapy, the patient was included in the study and considered to not have attained SVR. Secondary outcomes included a comparison of baseline characteristics and demographics between the 2 main cohorts and comparison between hospitalized patients who did and those who did not achieve SVR. Patients’ EMRs were reviewed to collect data for demographic, medical history, laboratory results, HCV treatment history and plan, and admissions within the health system. Hospital admissions were considered to be liver related if the admission reason was directly related to liver disease such as ascites, hepatic encephalopathy, or variceal bleed. Admission was not considered to be liver related if a secondary complication was involved—for example, acute kidney injury secondary to hypovolemia from diuretics used to manage ascites.
Sample Size Calculation and Statistical Methods
For sample size calculation and power analysis we utilized an expected SVR rate of 75% for patients hospitalized during treatment based on findings by Hill et al. 13 The control group was hypothesized to achieve an SVR rate of 95% per multiple DAA trials.1-8 Utilizing treatment expected outcomes of 95% and 75% with an α of 0.05 and 90% power, 75 patients per group were required for a total of 150 patients. Through preliminary reports and screening, we initially identified 138 potential patients admitted to UC Health facilities during treatment with a DAA for HCV. To create the control cohort, we hypothesized that up to 40% of patients prescribed an outpatient DAA regimen may not actually receive the medication because of numerous barriers. To obtain a 1:1 ratio of hospitalized to not hospitalized patients, and if all 138 potential hospitalized patients were included, we would need approximately 240 patients who were not hospitalized prior to exclusions. We obtained a report of all outpatient prescriptions for a DAA, which revealed 3750 prescriptions; from this report, 238 patients were randomly selected for the control cohort.
Baseline characteristics were compared using the t-test and Mann-Whitney rank sum test for continuous variables; χ2 and Fisher exact tests were used for analysis of categorical variables. Descriptive statistics were expressed in terms of medians unless otherwise specified. P values <0.05 were considered significant. We evaluated predictors of not achieving SVR in the overall cohort using a multivariate logistic regression model including risk factors that had P values <0.2 in the univariate analysis. Statistical analysis was performed using SigmaPlot version 14 (Systat Software; San Jose, CA).
Results
This study included 167 patients who were not hospitalized (30% not included, compared with hypothesized 40%, because patients had a prescription for a DAA but did not actually receive the medication) during DAA treatment for HCV and 94 patients who were admitted during DAA treatment. Baseline characteristics of patients are shown in Table 1. Patients in the hospitalized cohort were older and had more biopsy-confirmed or imaging-suggested cirrhosis, higher baseline Fibrosis-4 (FIB-4) score, higher Model for End-Stage Liver Disease (MELD) score, more severe Child Turcotte-Pugh (CTP) class, lower albumin and platelets, and higher international normalized ratio (INR); there was an increased number of recipients of a liver transplant, the patients had longer DAA treatment duration, and fewer patients were treatment naïve. Patients were similar in regard to sex, body mass index, race, HIV status or hepatitis B virus (HBV) coinfection, HCV genotype, and DAA treatment regimen. For the primary outcome, patients who were hospitalized during DAA treatment for HCV were found to have a significant reduction in attainment of SVR compared with those not hospitalized during treatment (87.1% vs 95.2%; P = 0.043). A secondary analysis was completed of the primary outcome, and no difference was found when excluding inpatient mortality in the hospitalized group (91.1% vs 95.2%; P = 0.195). All results retain full study population unless stated otherwise. Table 2 compares characteristics of hospitalized patients who achieved SVR versus those who did not achieve SVR. Patients were similar in regard to demographics, HIV or HBV coinfection, HCV genotype, cirrhosis, DAA regimen, new drug interactions during admission, and number of missed doses as documented on medication administration record (MAR). Patients who did achieve SVR were more likely to have had a liver transplant. Patients who were hospitalized and did not attain SVR had higher MELD-Na score on admission and were more likely to have an intensive care unit stay, have a longer length of stay, and to have inpatient mortality (33.3% vs 0%; P < 0.001). Of 94 hospitalized patients, 93 (98.9%) were able to supply their home DAAs for inpatient use at some point during admission. In the multivariate model designed to predict failure to achieve SVR, treatment duration, hospitalization, age, viral load, and CTP class A were included utilizing prespecified criteria (Table 3), but all failed to reach significance. The rate of SVR by number of missed doses is displayed in Figure 1 (figure excludes patients who died during admission), with patients missing 10 or more doses during hospital admission having the lowest rate of SVR (80.0%). Of the 4 hospitalized patients who died during admission (Table 4), all had prior liver transplant and had HCV genotype 1a or 1b. Two were admitted for direct liver-related complications. Two had high baseline MELD scores of 27 and 43 at the start of therapy, which were similar at hospital admission, with scores of 29 and 41. One patient experienced a large increase in MELD score from 15 at start of therapy to 28 at hospital admission.
Baseline Characteristics.
Abbreviations: CTP, Child-Turquotte-Pugh; FIB-4, Fibrosis-4; HBV, hepatitis B virus; HCV, hepatitis C virus; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; SCr, serum creatinine.
This does not include the regimen of sofosbuvir + ribavirin.
Hospitalized Patient Characteristics by Those Achieving and Not Achieving SVR.
Abbreviations: CTP, Child-Turcotte-Pugh; FIB-4, Fibrosis-4; HBV, hepatitis B virus; HCV, hepatitis C virus; ICU, intensive care unit; INR, international normalized ratio; IQR, interquartile range; MAR, medication administration record; MELD, Model for End-Stage Liver Disease; SCr, serum creatinine; SVR, sustained virologic response.
This does not include the regimen of sofosbuvir + ribavirin.
Multivariate Model to Predict Failure to Achieve SVR.
Abbreviations: CTP, Child Turcotte-Pugh; SVR, sustained virologic response.
Sustained virologic response (SVR) by number of missed inpatient doses.a
Inpatient Mortality.
Abbreviations: FIB-4, fibrosis-4; HBV, hepatitis B virus; HCV, hepatitis C virus; ICU, intensive care unit; MAR, medication administration record; MELD, Model for End-Stage Liver Disease; txp, transplant.
Discussion
Patients who were hospitalized during DAA treatment for HCV experienced less, although not statistically significant, treatment success than those not hospitalized during treatment when patients who died during admission were removed from analysis. Patients who were not hospitalized during treatment achieved similar SVR rates as those in controlled trials. Because achievement of SVR is associated with favorable clinical outcomes and DAA treatment cost is high, it is important to identify risk factors for treatment failure. This study’s objective was to evaluate if hospitalization during treatment affected SVR outcome and if specific risk factors during hospitalization such as DDIs or missed doses had influence.
DAAs have many DDIs and may be the causative agent or recipient of these interactions. 23 Common DDI pathways include cytochrome P450 or transporters such as p-glycoprotein, breast cancer resistance protein, or organic anion transporting polypeptide. Up to 80% of patients taking DAAs may be taking another medication that interacts; however, having a DDI alone has not been shown to affect SVR outcomes.24-26 One of the most common interactions with DAAs are proton pump inhibitors (PPIs), which are one of the most commonly prescribed medications worldwide. PPIs may reduce the bioavailability of DAAs, thereby reducing absorbed concentration. This led to package insert recommendations of a maximum PPI dose with concomitant use of certain DAAs.27-29 Despite this maximum dose suggestion, studies evaluating the significance of the PPI-DAA interaction have had conflicting results.30-32 Researchers evaluating sofosbuvir/ledipasvir for effects of the PPI DDI found no difference in SVR outcome between patients taking and those not taking a PPI; this cohort consisted of approximately 2000 total patients. 30 However, one research group evaluated 9 studies (>32 000 patients) and found that presence of a PPI reduced the odds of achieving SVR (OR = 0.74). 31 A majority of the patients were receiving sofosbuvir/ledipasvir. In our study, patients admitted during DAA treatment had high likelihood of receiving a PPI concomitant with a DAA. It is possible that acute care providers may not be as familiar with DDI for medications such as DAAs, which are largely considered outpatient medications. In our cohort, patients who received a PPI (that was not a home medication) during admission trended toward reduced SVR. Although this difference was not statistically significant, further evaluation of the PPI DDI in a larger cohort of hospitalized patients is warranted because our population was not powered to evaluate this. HCV treatment duration may be as short as 8 weeks, so it is vital that acute care providers monitor for DDIs because it is possible that a short-duration DDI may affect the treatment outcome of an 8- or 12-week DAA regimen.
It is also possible that missed doses during a short treatment course may adversely affect treatment outcome. A pooled analysis of 7 phase III clinical trials utilizing glecaprevir/pibrentasvir for 8 or 12 weeks compared outcomes among patients who were recent drug users, were former drug users, and had no prior drug use. 32 Treatment adherence was monitored by pill counts, and nonadherence (less than 90% by tablet count) was an independent predictor of not achieving SVR. Therefore, it is possible that missing as few as 5 or 6 doses of an 8-week regimen could negatively affect treatment outcome. In the hospitalized cohort of our study, 50% of hospitalized patients missed at least 1 dose per the MAR; however, the mean number of missed doses was similar between patients achieving or not achieving SVR. When evaluating SVR by number of missed doses, patients who missed ≥10 doses had only an 80.0% rate of achieving SVR (Figure 1). This reveals that a high number of missed doses may contribute to decreased treatment success. Because no definitive data exist regarding specific number of missed doses and lack of achieving SVR, some theorize that it is possible that a small number of missed doses may not have an effect on treatment outcomes because of DAA pharmacokinetic properties. For example, sofosbuvir has a short half-life of 0.5 hours, but the active metabolite of sofosbuvir (GS-331007) has a much longer half-life of 25 hours; velpatasvir has a half-life of 15 hours. 28 This may explain why a low mean number of missed doses appeared to have little effect in our cohort; however, our study may not have been large enough to statistically evaluate this difference in missed doses. Although the mean number of missed doses was similar between those who did and did not attain SVR in the hospitalized cohort, patients who missed 1 to 4 doses during admission had an SVR rate of 88.2% compared with 93.5% in patients who did not miss any doses during admission (P = 0.450; Figure 1). Although this was not statistically significant, it is possible that our population was not large enough to evaluate a small number of missed doses and studies with larger populations are needed. It was necessary to evaluate missed doses during hospitalization because HCV DAAs are often not included on hospital formularies because these medications are considered for outpatient treatment and are expensive to maintain stock with likely little utilization. The average length of stay for all hospitalized patients was 8 days, but the average number of missed doses was only 2. A likely reason for the low number of missed doses in this study may be a result of the high emphasis on patient education regarding policies of our health system to bring home supply of DAAs. Of all the hospitalized patients, only 1 did not utilize a home medication supply. It is not currently defined with certainty how many missed doses would negatively affect treatment outcome. Therefore, it is essential that patients are educated to bring home supplies of HCV DAAs if admitted to the hospital. Because of the high cost of HCV DAA regimens, we suggest that it is not necessary to add these medications to an institutional formulary when patients may bring in home supply. In a review of our hospitalized cohort, we found that patient education to bring in DAA supply is an effective intervention (98.9% provided home supply) to avoid unnecessary cost and reduce missed doses. However, it is possible that if a patient is expected to miss a high number of doses (Figure 1) because of a prolonged hospital stay and the patient is unable to provide the supply, other strategies to obtain medication may need to be pursued.
It is well recognized that patients with decompensated liver disease have reduced SVR rates even with the most current therapies. A phase 2 study evaluating sofosbuvir/ledipasvir plus ribavirin for HCV genotype 1 or 4 included patients with decompensated cirrhosis. 16 SVR rates for patients with CTP B were 87% receiving 12 weeks of treatment and 89% receiving 24 weeks of treatment. Patients with CTP C had SVR rates of 86% receiving 12 weeks and 87% receiving 24 weeks of therapy. Multiple real-world studies reporting sofosbuvir/ledipasvir use in decompensated cirrhosis have revealed SVR rates ranging from 81% to 90%.17,33 A phase 3 study evaluated sofosbuvir/velpatasvir for genotypes 1, 2, 3, 4, or 6 in decompensated cirrhosis with and without ribavirin and for 12- and 24-week treatment durations. 18 SVR rates ranged from 50% to 100% depending on genotype and whether the patient received ribavirin or not. Overall, our case cohort of patients had severe liver disease, with about 50% progressing to cirrhosis. The reduced treatment success of hospitalized patients seems likely to have been rooted in patients having more severe liver disease than those in the control cohort. When evaluating patients who were hospitalized, more had confirmed cirrhosis, higher FIB-4 and MELD scores, and prior liver transplant. Furthermore, 13.8% of patients not hospitalized had CTP B or C compared with 30.8% of hospitalized patients. One conflicting finding with this hypothesis, that liver disease led to reduced SVR, is the rates of liver-related hospitalizations (24.4% with SVR vs 33.3% with no SVR; P = 0.494), although it is possible that the study population was not large enough to evaluate this difference in cause of admission. In hospitalized patients only, 28.4% attaining SVR had CTP B or C compared with 50% not attaining SVR. Our findings correlate with previous findings that patients with more severe liver disease have reduced SVR. Although our study did not find a statistical difference in our secondary analysis that excluded mortality, the finding of 91.1% versus 95.2% SVR in hospitalized versus nonhospitalized patients should be explored further with a larger population.
In our cohort, 56.4% of hospitalized patients were treatment naïve versus 72.4% of nonhospitalized patients, with a significant difference (P = 0.018). This difference in prior treatment, in addition to more severe liver disease in the hospitalized group, led to the longer planned treatment duration in the hospitalized group because patients with treatment failure (and decompensated liver disease) require longer treatment duration depending on the medications used prior to and for retreatment. Prior treatment failure has been shown to influence SVR outcome depending on the medications used in the prior treatment and which medications were used in the retreatment. The current suggestion for DAA treatment-experienced patients is the combination of sofosbuvir/velpatasvir/voxilaprevir. Although successful (>95%) in clinical trials, it has only been approved by the Food and Drug Administration since 2017. In our cohort, none of the hospitalized patients received this medication, which could also affect SVR outcome because almost half had prior treatment.
A notable aspect of the hospitalized cohort is the marked increase in MELD score from baseline to hospital admission in patients who did not achieve SVR (12.5 to 17) compared with patients who did achieve SVR (10 to 10), who had a stable MELD score. Because hospitalized patients with a large increase in MELD scores from baseline to admission had reduced SVR, it may be necessary to explore strategies to prevent acute decompensations and evaluate care delivery factors prior to admission and during hospital stay. The AASLD recently published quality measures for patients with cirrhosis. 34 It may be a worthwhile investment to evaluate these quality measures specifically in patients receiving DAAs because improving quality may reduce hospitalization and improve care during hospitalization.
Our study has several limitations. First, the retrospective nature of the study prevents a conclusion of the direct effect of hospitalization on SVR outcome. Our hospitalized patients had reduced treatment success; however, delineating the causality of worsened outcomes with hospitalization, versus no hospitalization serving as a marker of at-risk patients, needs further exploration, and larger populations need to be evaluated. Patients in the hospitalized cohort had worse liver disease, which likely contributed to the reduced outcome. Ideally, to evaluate hospitalization influence on SVR, both cohorts would have similar baseline disease, although through our findings, it is likely that patients who are hospitalized are more likely to have more severe disease. We chose to include patients who had inpatient mortality in our primary analysis. This decision was made in order to provide the full picture of patients who are hospitalized and receiving DAAs in this real-life cohort. It is important to know how many patients died prior to completing therapy, and this provides further insights into patients who are hospitalized while receiving DAA treatment. Through their inclusion, we were able to observe a 4.2% mortality rate of admitted patients receiving a DAA. Including these patients in our primary analysis did not detract from the goal of our study because we were able to conduct a secondary analysis excluding patients who died during hospitalization. Another limitation is that we may not have identified all patients who were admitted and were receiving DAA therapy. Patients may have been admitted to a hospital outside of our health system. We used 1 year from time of original prescription because it is likely that patients will have delays in receiving the medication as a result of third-party payor eligibility and approval. However, a longer time period beyond 1 year would not have captured more patients because prescriptions expire beyond 1 year in our state. During hospital stay, we were able to quantify missed doses, but we were unable to evaluate medication adherence beyond the hospital stay in the outpatient setting. It is also possible that the patient could have been taking their home medication during hospital admission but the medication order was not entered into the EMR and, therefore, not documented as administered on the MAR, not allowing us to capture this patient. However, we ran a separate report of all outpatient DAA prescriptions of patients admitted within 1 year, and we could have captured the patient through this separate report. DAA regimens have changed over time, and we included regimens that are no longer commonly in use (eg, sofosbuvir plus simeprevir). However, the cohorts were balanced well with similar regimens used in each cohort. In addition, the vast majority of patients received regimens still currently in use.
Conclusion and Relevance
Patients receiving a DAA for HCV had reduced SVR outcomes, although not statistically significant when inpatient mortality was excluded, when hospitalized during DAA treatment when compared with those not hospitalized during treatment. Patients not achieving SVR had a higher degree of baseline liver dysfunction and experienced acute worsening on hospitalization, including death, which appears to be the most likely explanation for the reduced SVR. Patient education to provide home supply of DAAs during admission appears to be an effective intervention because these medications are not commonly included in hospital formularies. It may be important to evaluate care delivery factors prior to and during hospital admission especially in those with severe liver disease.
Footnotes
Declaration of Conflicting Interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: AJG served on a joint advisory board for Theravance BioPharma US, Inc, and Mylan Specialty LP. KES is the site principal investigator for clinical trials funded by AbbVie, Gilead, Merck, and Intercept, contracted through UC Health and the University of Cincinnati. He serves on the Data Safety Monitoring Boards for MedPace, Inovio, and Watermark.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
