Survival biases lead to flawed conclusions in observational treatment studies of influenza patients

Methodological biases in observational hospital studies of COVID-19 treatment effectiveness: pitfalls and potential

Introduction

This study aims to discuss and assess the impact of three prevalent methodological biases: competing risks, immortal-time bias, and confounding bias in real-world observational studies evaluating treatment effectiveness. We use a demonstrative observational data example of COVID-19 patients to assess the impact of these biases and propose potential solutions.

Methods

We describe competing risks, immortal-time bias, and time-fixed confounding bias by evaluating treatment effectiveness in hospitalized patients with COVID-19. For our demonstrative analysis, we use observational data from the registry of patients with COVID-19 who were admitted to the Bellvitge University Hospital in Spain from March 2020 to February 2021 and met our predefined inclusion criteria. We compare estimates of a single-dose, time-dependent treatment with the standard of care. We analyze the treatment effectiveness using common statistical approaches, either by ignoring or only partially accounting for the methodological biases. To address these challenges, we emulate a target trial through the clone-censor-weight approach.

Results

Overlooking competing risk bias and employing the naïve Kaplan-Meier estimator led to increased in-hospital death probabilities in patients with COVID-19. Specifically, in the treatment effectiveness analysis, the Kaplan-Meier estimator resulted in an in-hospital mortality of 45.6% for treated patients and 59.0% for untreated patients. In contrast, employing an emulated trial framework with the weighted Aalen-Johansen estimator, we observed that in-hospital death probabilities were reduced to 27.9% in the "X"-treated arm and 40.1% in the non-"X"-treated arm. Immortal-time bias led to an underestimated hazard ratio of treatment.

Conclusion

Overlooking competing risks, immortal-time bias, and confounding bias leads to shifted estimates of treatment effects. Applying the naïve Kaplan-Meier method resulted in the most biased results and overestimated probabilities for the primary outcome in analyses of hospital data from COVID-19 patients. This overestimation could mislead clinical decision-making. Both immortal-time bias and confounding bias must be addressed in assessments of treatment effectiveness. The trial emulation framework offers a potential solution to address all three methodological biases.

Target trial emulation with multi-state model analysis to assess treatment effectiveness using clinical COVID-19 data

BACKGROUND

Real-world observational data are an important source of evidence on the treatment effectiveness for patients hospitalized with coronavirus disease 2019 (COVID-19). However, observational studies evaluating treatment effectiveness based on longitudinal data are often prone to methodological biases such as immortal time bias, confounding bias, and competing risks.

METHODS

For exemplary target trial emulation, we used a cohort of patients hospitalized with COVID-19 (n = 501) in a single centre. We described the methodology for evaluating the effectiveness of a single-dose treatment, emulated a trial using real-world data, and drafted a hypothetical study protocol describing the main components. To avoid immortal time and time-fixed confounding biases, we applied the clone-censor-weight technique. We set a 5-day grace period as a period of time when treatment could be initiated. We used the inverse probability of censoring weights to account for the selection bias introduced by artificial censoring. To estimate the treatment effects, we took the multi-state model approach. We considered a multi-state model with five states. The primary endpoint was defined as clinical severity status, assessed by a 5-point ordinal scale on day 30. Differences between the treatment group and standard of care treatment group were calculated using a proportional odds model and shown as odds ratios. Additionally, the weighted cause-specific hazards and transition probabilities for each treatment arm were presented.

RESULTS

Our study demonstrates that trial emulation with a multi-state model analysis is a suitable approach to address observational data limitations, evaluate treatment effects on clinically heterogeneous in-hospital death and discharge alive endpoints, and consider the intermediate state of admission to ICU. The multi-state model analysis allows us to summarize results using stacked probability plots that make it easier to interpret results.

CONCLUSIONS

Extending the emulated target trial approach to multi-state model analysis complements treatment effectiveness analysis by gaining information on competing events. Combining two methodologies offers an option to address immortal time bias, confounding bias, and competing risk events. This methodological approach can provide additional insight for decision-making, particularly when data from randomized controlled trials (RCTs) are unavailable.

Bias in observational studies on the effectiveness of in hospital use of hydroxychloroquine in COVID-19

PURPOSE

During the first waves of the coronavirus pandemic, evidence on potential effective treatments was urgently needed. Results from observational studies on the effectiveness of hydroxychloroquine (HCQ) were conflicting, potentially due to biases. We aimed to assess the quality of observational studies on HCQ and its relation to effect sizes.

METHODS

PubMed was searched on 15 March 2021 for observational studies on the effectiveness of in-hospital use of HCQ in COVID-19 patients, published between 01/01/2020 and 01/03/2021 on. Study quality was assessed using the ROBINS-I tool. Association between study quality and study characteristics (journal ranking, publication date, and time between submission and publication) and differences between effects sizes found in observational studies compared to those found in RCTs, were assessed using Spearman's correlation.

RESULTS

Eighteen of the 33 (55%) included observational studies were scored as critical risk of bias, eleven (33%) as serious risk and only four (12%) as moderate risk of bias. Biases were most often scored as critical in the domains related to selection of participants (n = 13, 39%) and bias due to confounding (n = 8, 24%). There were no significant associations found between the study quality and the characteristics nor between the study quality and the effect estimates.

DISCUSSION

Overall, the quality of observational HCQ studies was heterogeneous. Synthesis of evidence of effectiveness of HCQ in COVID-19 should focus on RCTs and carefully consider the added value and quality of observational evidence.

Target Trial Emulation Using Hospital-Based Observational Data: Demonstration and Application in COVID-19

Methodological biases are common in observational studies evaluating treatment effectiveness. The objective of this study is to emulate a target trial in a competing risks setting using hospital-based observational data. We extend established methodology accounting for immortal time bias and time-fixed confounding biases to a setting where no survival information beyond hospital discharge is available: a condition common to coronavirus disease 2019 (COVID-19) research data. This exemplary study includes a cohort of 618 hospitalized patients with COVID-19. We describe methodological opportunities and challenges that cannot be overcome applying traditional statistical methods. We demonstrate the practical implementation of this trial emulation approach via clone-censor-weight techniques. We undertake a competing risk analysis, reporting the cause-specific cumulative hazards and cumulative incidence probabilities. Our analysis demonstrates that a target trial emulation framework can be extended to account for competing risks in COVID-19 hospital studies. In our analysis, we avoid immortal time bias, time-fixed confounding bias, and competing risks bias simultaneously. Choosing the length of the grace period is justified from a clinical perspective and has an important advantage in ensuring reliable results. This extended trial emulation with the competing risk analysis enables an unbiased estimation of treatment effects, along with the ability to interpret the effectiveness of treatment on all clinically important outcomes.

Concurrent analysis of hospital stay durations and mortality of emerging severe acute respiratory coronavirus virus 2 (SARS-CoV-2) variants using real-time electronic health record data at a large German university hospital

Multistate methodology proves effective in analyzing hospitalized coronavirus disease 2019 (COVID-19) patients with emerging variants in real time. An analysis of 2,548 admissions in Freiburg, Germany, showed reduced severity over time in terms of shorter hospital stays and higher discharge rates when comparing more recent phases with earlier phases of the pandemic.

Management of Severe Influenza

Influenza infection causes severe illness in 3 to 5 million people annually, with up to an estimated 650,000 deaths per annum. As such, it represents an ongoing burden to health care systems and human health. Severe acute respiratory infection can occur, resulting in respiratory failure requiring intensive care support. Herein we discuss diagnostic approaches, including development of CLIA-waived point of care tests that allow rapid diagnosis and treatment of influenza. Bacterial and fungal coinfections in severe influenza pneumonia are associated with worse outcomes, and we summarize the approach and treatment options for diagnosis and treatment of bacterial and Aspergillus coinfection. We discuss the available drug options for the treatment of severe influenza, and treatments which are no longer supported by the evidence base. Finally, we describe the supportive management and ventilatory approach to patients with respiratory failure as a result of severe influenza in the intensive care unit.

Enani M, Saheb Sharif-Askari F, R. Alghareeb M, T. Bin-Brikan R, A. AlShahrani S, Halwani R, Tleyjeh IM. Combination of (interferon beta-1b, lopinavir/ritonavir and ribavirin) versus favipiravir in hospitalized patients with non-critical COVID-19: A cohort study

Objectives

Our study aims at comparing the efficacy and safety of IFN-based therapy (lopinavir/ritonavir, ribavirin, and interferon β-1b) vs. favipiravir (FPV) in a cohort of hospitalized patients with non-critical COVID-19.

Methods

Single center observational study comparing IFN-based therapy (interferon β-1b, ribavirin, and lopinavir/ritonavir) vs. FPV in non-critical hospitalized COVID-19 patients. Allocation to either treatment group was non-random but based on changes to national treatment protocols rather than physicians’ selection (quasi-experimental). We examined the association between IFN-based therapy and 28-day mortality using Cox regression model with treatment as a time-dependent covariate.

Results

The study cohort included 222 patients, of whom 68 (28%) received IFN-based therapy. Antiviral therapy was started at a median of 5 days (3–6 days) from symptoms onset in the IFN group vs. 6 days (4–7 days) for the FPV group, P <0.0001. IFN-based therapy was associated with a lower 28-day mortality as compared to FPV (6 (9%) vs. 18 (12%)), adjusted hazard ratio [aHR] (95% Cl) = 0.27 (0.08–0.88)). No difference in hospitalization duration between the 2 groups, 9 (7–14) days vs. 9 (7–13) days, P = 0.732 was found. IFN treated group required less use of systemic corticosteroids (57%) as compared to FPV (77%), P = 0.005 after adjusting for disease severity and other confounders. Patients in the IFN treated group were more likely to have nausea and diarrhea as compared to FPV group (13%) vs. (3%), P = 0.013 and (18%) vs. (3%), P<0.0001, respectively.

Conclusion

Early IFN-based triple therapy was associated with lower 28-days mortality as compared to FPV.

Influenza Therapeutics in Clinical Practice-Challenges and Recent Advances

In the last few years, several new direct-acting influenza antivirals have been licensed, and others have advanced in clinical development. The increasing diversity of antiviral classes should allow an adequate public health response should a resistant virus to one agent or class widely circulate. One new antiviral, baloxavir marboxil, has been approved in the United States for treatment of influenza in those at high risk of developing influenza-related complications. Except for intravenous zanamivir in European Union countries, no antivirals have been licensed specifically for the indication of severe influenza or hospitalized influenza. This review addresses recent clinical developments involving selected polymerase inhibitors, neuraminidase inhibitors, antibody-based therapeutics, and host-directed therapies. There are many knowledge gaps for most of these agents because some data are not published and multiple pivotal studies are in progress at present. This review also considers important clinical research issues, including regulatory pathways, study designs, endpoints, and target populations encountered during the clinical development of novel therapeutics.

Methodological evaluation of bias in observational coronavirus disease 2019 studies on drug effectiveness

Background and objective

Observational studies may provide valuable evidence on real-world causal effects of drug effectiveness in patients with coronavirus disease 2019 (COVID-19). As patients are usually observed from hospital admission to discharge and drug initiation starts during hospitalization, advanced statistical methods are needed to account for time-dependent drug exposure, confounding and competing events. Our objective is to evaluate the observational studies on the three common methodological pitfalls in time-to-event analyses: immortal time bias, confounding bias and competing risk bias.

Methods

We performed a systematic literature search on 23 October 2020, in the PubMed database to identify observational cohort studies that evaluated drug effectiveness in hospitalized patients with COVID-19. We included articles published in four journals: British Medical Journal, New England Journal of Medicine, Journal of the American Medical Association and The Lancet as well as their sub-journals.

Results

Overall, out of 255 articles screened, 11 observational cohort studies on treatment effectiveness with drug exposure–outcome associations were evaluated. All studies were susceptible to one or more types of bias in the primary study analysis. Eight studies had a time-dependent treatment. However, the hazard ratios were not adjusted for immortal time in the primary analysis. Even though confounders presented at baseline have been addressed in nine studies, time-varying confounding caused by time-varying treatment exposure and clinical variables was less recognized. Only one out of 11 studies addressed competing event bias by extending follow-up beyond patient discharge.

Conclusions

In the observational cohort studies on drug effectiveness for treatment of COVID-19 published in four high-impact journals, the methodological biases were concerningly common. Appropriate statistical tools are essential to avoid misleading conclusions and to obtain a better understanding of potential treatment effects.

Early oseltamivir treatment improves survival in critically ill patients with influenza pneumonia

Background

The relationship between early oseltamivir treatment (within 48 h of symptom onset) and mortality in patients admitted to intensive care units (ICUs) with severe influenza is disputed. This study aimed to investigate the association between early oseltamivir treatment and ICU mortality in critically ill patients with influenza pneumonia.

Methods

This was an observational study of patients with influenza pneumonia admitted to 184 ICUs in Spain during 2009-2018. The primary outcome was to evaluate the association between early oseltamivir treatment and ICU mortality compared with later treatment. Secondary outcomes were to compare the duration of mechanical ventilation and ICU length of stay between the early and later oseltamivir treatment groups. To reduce biases related to observational studies, propensity score matching and a competing risk analysis were performed.

Results

During the study period, 2124 patients met the inclusion criteria. All patients had influenza pneumonia and received oseltamivir before ICU admission. Of these, 529 (24.9%) received early oseltamivir treatment. In the multivariate analysis, early treatment was associated with reduced ICU mortality (OR 0.69, 95% CI 0.51-0.95). After propensity score matching, early oseltamivir treatment was associated with improved survival rates in the Cox regression (hazard ratio 0.77, 95% CI 0.61-0.99) and competing risk (subdistribution hazard ratio 0.67, 95% CI 0.53-0.85) analyses. The ICU length of stay and duration of mechanical ventilation were shorter in patients receiving early treatment.

Conclusions

Early oseltamivir treatment is associated with improved survival rates in critically ill patients with influenza pneumonia, and may decrease ICU length of stay and mechanical ventilation duration.

Efficacy and safety of tocilizumab in COVID-19 patients: a living systematic review and meta-analysis

OBJECTIVES

Cytokine release syndrome with elevated interleukin-6 (IL-6) levels is associated with multiorgan damage and death in severe coronavirus disease 2019 (COVID-19). Our objective was to perform a living systematic review of the literature concerning the efficacy and toxicity of the IL-6 receptor antagonist tocilizumab in COVID-19 patients.

METHODS

Data sources were Ovid MEDLINE(R) and Epub Ahead of Print, In-Process & Other Non-Indexed Citations and Daily, Ovid Embase, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, Web of Science, Scopus up, preprint servers and Google up to October 8, 2020. Study eligibility criteria were randomized controlled trials (RCTs) and observational studies at low or moderate risk of bias. Participants were hospitalized COVID-19 patients. Interventions included tocilizumab versus placebo or standard of care. We pooled crude risk ratios (RRs) of RCTs and adjusted RRs from cohorts, separately. We evaluated inconsistency between studies with I2. We assessed the certainty of evidence using the GRADE approach.

RESULTS

Of 1156 citations, 24 studies were eligible (five RCTs and 19 cohorts). Five RCTs at low risk of bias, with 1325 patients, examined the effect of tocilizumab on short-term mortality; pooled RR was 1.09 (95%CI 0.80-1.49, I2 = 0%). Four RCTs with 771 patients examined the effect of tocilizumab on risk of mechanical ventilation; pooled RR was 0.71 (95%CI 0.52-0.96, I2 = 0%), with a corresponding number needed to treat of 17 (95%CI 9-100). Among 18 cohorts at moderate risk of bias with 9850 patients, the pooled adjusted RR for mortality was 0.58 (95%CI 0.51-0.66, I2 = 2.5%). This association was observed over all degrees of COVID-19 severity. Data from the RCTs did not show a higher risk of infections or adverse events with tocilizumab: pooled RR 0.63 (95%CI 0.38-1.06, five RCTs) and 0.83 (95%CI 0.55-1.24, five RCTs), respectively.

CONCLUSIONS

Cumulative moderate-certainty evidence shows that tocilizumab reduces the risk of mechanical ventilation in hospitalized COVID-19 patients. While RCTs showed that tocilizumab did not reduce short-term mortality, low-certainty evidence from cohort studies suggests an association between tocilizumab and lower mortality. We did not observe a higher risk of infections or adverse events with tocilizumab use. This review will continuously evaluate the role of tocilizumab in COVID-19 treatment.

Efficacy of chloroquine or hydroxychloroquine in COVID-19 patients: a systematic review and meta-analysis

OBJECTIVES

Clinical studies of chloroquine (CQ) and hydroxychloroquine (HCQ) in COVID-19 disease reported conflicting results. We sought to systematically evaluate the effect of CQ and HCQ with or without azithromycin on outcomes of COVID-19 patients.

METHODS

We searched multiple databases, preprints and grey literature up to 17 July 2020. We pooled only adjusted-effect estimates of mortality using a random-effect model. We summarized the effect of CQ or HCQ on viral clearance, ICU admission/mechanical ventilation and hospitalization.

RESULTS

Seven randomized clinical trials (RCTs) and 14 cohort studies were included (20 979 patients). Thirteen studies (1 RCT and 12 cohort studies) with 15 938 hospitalized patients examined the effect of HCQ on short-term mortality. The pooled adjusted OR was 1.05 (95% CI 0.96-1.15, I2 = 0%). Six cohort studies examined the effect of the HCQ+azithromycin combination with a pooled adjusted OR of 1.32 (95% CI 1.00-1.75, I2 = 68.1%). Two cohort studies and four RCTs found no effect of HCQ on viral clearance. One small RCT demonstrated improved viral clearance with CQ and HCQ. Three cohort studies found that HCQ had no significant effect on mechanical ventilation/ICU admission. Two RCTs found no effect for HCQ on hospitalization risk in outpatients with COVID-19.

CONCLUSIONS

Moderate certainty evidence suggests that HCQ, with or without azithromycin, lacks efficacy in reducing short-term mortality in patients hospitalized with COVID-19 or risk of hospitalization in outpatients with COVID-19.

Association of corticosteroids use and outcomes in COVID-19 patients: A systematic review and meta-analysis

BACKGROUND

To systematically review the literature about the association between systemic corticosteroid therapy (CST) and outcomes of COVID-19 patients.

METHODS

We searched Medline, Embase, EBM Reviews, Scopus, Web of Science, and preprints up to July 20, 2020. We included observational studies and randomized controlled trials (RCT) that assessed COVID-19 patients treated with CST. We pooled adjusted effect estimates of mortality and other outcomes using a random effect model, among studies at low or moderate risk for bias. We assessed the certainty of evidence for each outcome using the GRADE approach.

RESULTS

Out of 1067 citations screened for eligibility, one RCT and 19 cohort studies were included (16,977 hospitalized patients). Ten studies (1 RCT and 9 cohorts) with 10,278 patients examined the effect of CST on short term mortality. The pooled adjusted RR was 0.92 (95% CI 0.69-1.22, I2 = 81.94%). This effect was observed across all stages of disease severity. Four cohort studies examined the effect of CST on composite outcome of death, ICU admission and mechanical ventilation need. The pooled adjusted RR was 0.41(0.23-0.73, I2 = 78.69%). Six cohort studies examined the effect of CST on delayed viral clearance. The pooled adjusted RR was 1.47(95% CI 1.11-1.93, I2 = 43.38%).

CONCLUSION

In this systematic review, as of July 2020, heterogeneous and low certainty cumulative evidence based on observational studies and one RCT suggests that CST was not associated with reduction in short-term mortality but possibly with a delay in viral clearance in patients hospitalized with COVID-19 of different severities. However, the discordant results between the single RCT and observational studies as well as the heterogeneity observed across observational studies, call for caution in using observational data and suggests the need for more RCTs to identify the clinical and biochemical characteristics of patients' population that could benefit from CST.

Effect of Early Oseltamivir Treatment on Mortality in Critically Ill Patients With Different Types of Influenza: A Multiseason Cohort Study

BACKGROUND

The available evidence on whether neuraminidase inhibitors reduce mortality in patients with influenza is inconclusive and focuses solely on influenza A/H1N1pdm09. We assessed whether early oseltamivir treatment (≤48 hours from symptom onset) decreases mortality compared to late treatment in a large cohort of critically ill patients with influenza of all types.

METHODS

The study included all adults with laboratory-confirmed influenza hospitalized in intensive care units (ICUs) in Greece over 8 seasons (2010-2011 to 2017-2018) and treated with oseltamivir. The association of early oseltamivir with mortality was assessed with log-binomial models and a competing risks analysis estimating cause-specific and subdistribution hazards for death and discharge. Effect estimates were stratified by influenza type and adjusted for multiple covariates.

RESULTS

A total of 1330 patients were studied, of whom 622 (46.8%) died in the ICU. Among patients with influenza A/H3N2, early treatment was associated with significantly lower mortality (relative risk, 0.69 [95% credible interval {CrI}, .49-.94]; subdistribution hazard ratio, 0.58 [95% CrI, .37-.88]). This effect was purely due to an increased cause-specific hazard for discharge, whereas the cause-specific hazard for death was not increased. Among survivors, the median length of ICU stay was shorter with early treatment by 1.8 days (95% CrI, .5-3.5 days). No effect on mortality was observed for A/H1N1 and influenza B patients.

CONCLUSIONS

Severely ill patients with suspected influenza should be promptly treated with oseltamivir, particularly when A/H3N2 is circulating. The efficacy of oseltamivir should not be assumed to be equal against all types of influenza.

Influenza-associated mortality in hospital care: a retrospective cohort study of risk factors and impact of oseltamivir in an English teaching hospital, 2016 to 2017

Background

Evidence of an oseltamivir treatment effect on influenza A(H3N2) virus infections in hospitalised patients is incomplete.

Aims

This cohort study aimed to evaluate risk factors for death among PCR-confirmed hospitalised cases of seasonal influenza A(H3N2) of all ages and the impact of oseltamivir.

Methods

Participants included all 332 PCR-confirmed influenza A(H3N2) cases diagnosed between 30 August 2016 and 17 March 2017 in an English university teaching Hospital. Oseltamivir treatment effect on odds of inpatient death was assessed by backward stepwise multivariable logistic regression analysis.

Results

The odds of death were reduced by two thirds (odds ratio (OR): 0.32; 95% confidence interval (CI): 0.11–0.93), in inpatients treated with a standard course of oseltamivir 75 mg two times daily for 5 days – compared with those untreated with oseltamivir, after adjustment for age, sex, current excess alcohol intake, receipt of 2016/17 seasonal influenza vaccine, serum haemoglobin and hospital vs community attribution of acquisition of influenza.

Conclusions

Oseltamivir treatment given according to National Institutes of Clinical Excellence (NICE); United States Centres for Disease Control and Prevention (CDC); Infectious Diseases Society of America (IDSA) and World Health Organization (WHO) guidelines was shown to be effective in reducing the odds of mortality in inpatients with PCR-confirmed seasonal influenza A(H3N2) after adjustment in a busy routine English hospital setting. Our results highlight the importance of hospitals complying with relevant guidelines for prompt seasonal influenza PCR testing and ensuring standard oseltamivir treatment to all PCR-confirmed cases of seasonal influenza.

Statistical Analysis of Clinical COVID-19 Data: A Concise Overview of Lessons Learned, Common Errors and How to Avoid Them

By definition, in-hospital patient data are restricted to the time between hospital admission and discharge (alive or dead). For hospitalised cases of COVID-19, a number of events during hospitalization are of interest regarding the influence of risk factors on the likelihood of experiencing these events. The same is true for predicting times from hospital admission of COVID-19 patients to intensive care or from start of ventilation (invasive or non-invasive) to extubation. This logical restriction of the data to the period of hospitalisation is associated with a substantial risk that inappropriate methods are used for analysis. Here, we briefly discuss the most common types of bias which can occur when analysing in-hospital COVID-19 data.

Joint analysis of duration of ventilation, length of intensive care, and mortality of COVID-19 patients: a multistate approach

BACKGROUND

The clinical progress of patients hospitalized due to COVID-19 is often associated with severe pneumonia which may require intensive care, invasive ventilation, or extracorporeal membrane oxygenation (ECMO). The length of intensive care and the duration of these supportive therapies are clinically relevant outcomes. From the statistical perspective, these quantities are challenging to estimate due to episodes being time-dependent and potentially multiple, as well as being determined by the competing, terminal events of discharge alive and death.

METHODS

We used multistate models to study COVID-19 patients' time-dependent progress and provide a statistical framework to estimate hazard rates and transition probabilities. These estimates can then be used to quantify average sojourn times of clinically important states such as intensive care and invasive ventilation. We have made two real data sets of COVID-19 patients (n = 24* and n = 53**) and the corresponding statistical code publically available.

RESULTS

The expected lengths of intensive care unit (ICU) stay at day 28 for the two cohorts were 15.05* and 19.62** days, while expected durations of mechanical ventilation were 7.97* and 9.85** days. Predicted mortality stood at 51%* and 15%**. Patients mechanically ventilated at the start of the example studies had a longer expected duration of ventilation (12.25*, 14.57** days) compared to patients non-ventilated (4.34*, 1.41** days) after 28 days. Furthermore, initially ventilated patients had a higher risk of death (54%* and 20%** vs. 48%* and 6%**) after 4 weeks. These results are further illustrated in stacked probability plots for the two groups from time zero, as well as for the entire cohort which depicts the predicted proportions of the patients in each state over follow-up.

CONCLUSIONS

The multistate approach gives important insights into the progress of COVID-19 patients in terms of ventilation duration, length of ICU stay, and mortality. In addition to avoiding frequent pitfalls in survival analysis, the methodology enables active cases to be analyzed by allowing for censoring. The stacked probability plots provide extensive information in a concise manner that can be easily conveyed to decision makers regarding healthcare capacities. Furthermore, clear comparisons can be made among different baseline characteristics.