Fusing Randomized Trials With Big Data: The Key to Self-learning Health Care Systems?

Adaptation of the Risk Analysis Index for Frailty Assessment Using Diagnostic Codes

Importance

Frailty is associated with adverse outcomes after even minor physiologic stressors. The validated Risk Analysis Index (RAI) quantifies frailty; however, existing methods limit application to in-person interview (clinical RAI) and quality improvement datasets (administrative RAI).

Objective

To expand the utility of the RAI utility to available International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) administrative data, using the National Inpatient Sample (NIS).

Design, Setting, and Participants

RAI parameters were systematically adapted to ICD-10-CM codes (RAI-ICD) and were derived (NIS 2019) and validated (NIS 2020). The primary analysis included survey-weighed discharge data among adults undergoing major surgical procedures. Additional external validation occurred by including all operative and nonoperative hospitalizations in the NIS (2020) and in a multihospital health care system (UPMC, 2021-2022). Data analysis was conducted from January to May 2023.

Exposures

RAI parameters and in-hospital mortality.

Main Outcomes and Measures

The association of RAI parameters with in-hospital mortality was calculated and weighted using logistic regression, generating an integerized RAI-ICD score. After initial validation, thresholds defining categories of frailty were selected by a full complement of test statistics. Rates of elective admission, length of stay, hospital charges, and in-hospital mortality were compared across frailty categories. C statistics estimated model discrimination.

Results

RAI-ICD parameters were weighted in the 9 548 206 patients who were hospitalized (mean [SE] age, 55.4 (0.1) years; 3 742 330 male [weighted percentage, 39.2%] and 5 804 431 female [weighted percentage, 60.8%]), modeling in-hospital mortality (2.1%; 95% CI, 2.1%-2.2%) with excellent derivation discrimination (C statistic, 0.810; 95% CI, 0.808-0.813). The 11 RAI-ICD parameters were adapted to 323 ICD-10-CM codes. The operative validation population of 8 113 950 patients (mean [SE] age, 54.4 (0.1) years; 3 148 273 male [weighted percentage, 38.8%] and 4 965 737 female [weighted percentage, 61.2%]; in-hospital mortality, 2.5% [95% CI, 2.4%-2.5%]) mirrored the derivation population. In validation, the weighted and integerized RAI-ICD yielded good to excellent discrimination in the NIS operative sample (C statistic, 0.784; 95% CI, 0.782-0.786), NIS operative and nonoperative sample (C statistic, 0.778; 95% CI, 0.777-0.779), and the UPMC operative and nonoperative sample (C statistic, 0.860; 95% CI, 0.857-0.862). Thresholds defining robust (RAI-ICD <27), normal (RAI-ICD, 27-35), frail (RAI-ICD, 36-45), and very frail (RAI-ICD >45) strata of frailty maximized precision (F1 = 0.33) and sensitivity and specificity (Matthews correlation coefficient = 0.26). Adverse outcomes increased with increasing frailty.

Conclusion and Relevance

In this cohort study of hospitalized adults, the RAI-ICD was rigorously adapted, derived, and validated. These findings suggest that the RAI-ICD can extend the quantification of frailty to inpatient adult ICD-10-CM-coded patient care datasets.

The Use of Artificial Intelligence in Head and Neck Cancers: A Multidisciplinary Survey

Artificial intelligence (AI) approaches have been introduced in various disciplines but remain rather unused in head and neck (H&N) cancers. This survey aimed to infer the current applications of and attitudes toward AI in the multidisciplinary care of H&N cancers. From November 2020 to June 2022, a web-based questionnaire examining the relationship between AI usage and professionals' demographics and attitudes was delivered to different professionals involved in H&N cancers through social media and mailing lists. A total of 139 professionals completed the questionnaire. Only 49.7% of the respondents reported having experience with AI. The most frequent AI users were radiologists (66.2%). Significant predictors of AI use were primary specialty (V = 0.455; p < 0.001), academic qualification and age. AI's potential was seen in the improvement of diagnostic accuracy (72%), surgical planning (64.7%), treatment selection (57.6%), risk assessment (50.4%) and the prediction of complications (45.3%). Among participants, 42.7% had significant concerns over AI use, with the most frequent being the 'loss of control' (27.6%) and 'diagnostic errors' (57.0%). This survey reveals limited engagement with AI in multidisciplinary H&N cancer care, highlighting the need for broader implementation and further studies to explore its acceptance and benefits.

Effect of clinical decision support for severe hypercholesterolemia on low-density lipoprotein cholesterol levels

Severe hypercholesterolemia/possible familial hypercholesterolemia (FH) is relatively common but underdiagnosed and undertreated. We investigated whether implementing clinical decision support (CDS) was associated with lower low-density lipoprotein cholesterol (LDL-C) in patients with severe hypercholesterolemia/possible FH (LDL-C ≥ 190 mg/dL). As part of a pre-post implementation study, a CDS alert was deployed in the electronic health record (EHR) in a large health system comprising 3 main sites, 16 hospitals and 53 clinics. Data were collected for 3 months before ('silent mode') and after ('active mode') its implementation. Clinicians were only able to view the alert in the EHR during active mode. We matched individuals 1:1 in both modes, based on age, sex, and baseline lipid lowering therapy (LLT). The primary outcome was difference in LDL-C between the two groups and the secondary outcome was initiation/intensification of LLT after alert trigger. We identified 800 matched patients in each mode (mean ± SD age 56.1 ± 11.8 y vs. 55.9 ± 11.8 y; 36.0% male in both groups; mean ± SD initial LDL-C 211.3 ± 27.4 mg/dL vs. 209.8 ± 23.9 mg/dL; 11.2% on LLT at baseline in each group). LDL-C levels were 6.6 mg/dL lower (95% CI, -10.7 to -2.5; P = 0.002) in active vs. silent mode. The odds of high-intensity statin use (OR, 1.78; 95% CI, 1.41-2.23; P < 0.001) and LLT initiation/intensification (OR, 1.30, 95% CI, 1.06-1.58, P = 0.01) were higher in active vs. silent mode. Implementation of a CDS was associated with lowering of LDL-C levels in patients with severe hypercholesterolemia/possible FH, likely due to higher rates of clinician led LLT initiation/intensification.

Creating a Statistical Analysis Plan to Continually Evaluate Intervention Adaptations that Arise in Real-World Implementation

Evidence-based health interventions are frequently translated into real-world settings where practical needs drive changes to intervention protocols. Due to logistical and resource constraints, these naturally arising adaptations are rarely assessed for comparative effectiveness using a randomized trial. Nevertheless, when observational data are available, it is still possible to identify beneficial adaptations using statistical methods that adjust for differences among intervention groups. As implementation continues and more data are collected and assessed, we also require analysis methods that ensure low statistical error rates as multiple comparisons are made over time. This paper describes how to create a statistical analysis plan for evaluating adaptations to an intervention during ongoing implementation. This can be done by combining methods commonly used in platform clinical trials with methods used for real-world data. We also demonstrate how to use simulations based on previous data to decide the frequency with which to conduct statistical analyses. The illustration uses data from large-scale implementation of a school-based resilience and skill-building preventive intervention to which several adaptations were made. The proposed statistical analysis plan for evaluating the school-based intervention has potential to improve population-level outcomes as implementation scales up further and additional adaptations are anticipated.

A Machine Learning Approach to Assess Patients with Deep Neck Infection Progression to Descending Mediastinitis: Preliminary Results

BACKGROUND

Deep neck infection (DNI) is a serious infectious disease, and descending mediastinitis is a fatal infection of the mediastinum. However, no study has applied artificial intelligence to assess progression to descending mediastinitis in DNI patients. Thus, we developed a model to assess the possible progression of DNI to descending mediastinitis.

METHODS

Between August 2017 and December 2022, 380 patients with DNI were enrolled; 75% of patients (n = 285) were assigned to the training group for validation, whereas the remaining 25% (n = 95) were assigned to the test group to determine the accuracy. The patients' clinical and computed tomography (CT) parameters were analyzed via the k-nearest neighbor method. The predicted and actual progression of DNI patients to descending mediastinitis were compared.

RESULTS

In the training and test groups, there was no statistical significance (all p > 0.05) noted at clinical variables (age, gender, chief complaint period, white blood cells, C-reactive protein, diabetes mellitus, and blood sugar), deep neck space (parapharyngeal, submandibular, retropharyngeal, and multiple spaces involved, ≥3), tracheostomy performance, imaging parameters (maximum diameter of abscess and nearest distance from abscess to level of sternum notch), or progression to mediastinitis. The model had a predictive accuracy of 82.11% (78/95 patients), with sensitivity and specificity of 41.67% and 87.95%, respectively.

CONCLUSIONS

Our model can assess the progression of DNI to descending mediastinitis depending on clinical and imaging parameters. It can be used to identify DNI patients who will benefit from prompt treatment.

Treating COVID-19: Targeting the Host Response, Not the Virus

In low- and middle-income countries (LMICs), inexpensive generic drugs like statins, ACE inhibitors, and ARBs, especially if used in combination, might be the only practical way to save the lives of patients with severe COVID-19. These drugs will already be available in all countries on the first pandemic day. Because they target the host response to infection instead of the virus, they could be used to save lives during any pandemic. Observational studies show that inpatient statin treatment reduces 28–30-day mortality but randomized controlled trials have failed to show this benefit. Combination treatment has been tested for antivirals and dexamethasone but, with the exception of one observational study in Belgium, not for inexpensive generic drugs. Future pandemic research must include testing combination generic drug treatments that could be used in LMICs.

The Importance of Pragmatic Study Design to the Scholarly Influence of Surgical Hip Fracture Randomized Controlled Trials

BACKGROUND

Surgical randomized controlled trials (RCTs) have potential drawbacks, leading some to question their role in filling the information gap in orthopaedic surgery. Pragmatism in study design was introduced to increase the clinical applicability of study results. The purpose of this study was to examine how pragmatism affects the scholarly influence of surgical RCTs.

METHODS

A search for surgical hip fracture-related RCTs published between 1995 and 2015 was done. Journal impact factor, citation number, research question, significance and type of outcome, number of centers involved, and the Pragmatic-Explanatory Continuum Indicator Summary-2 level of pragmatism score were recorded for each study. Scholarly influence was estimated by a study's inclusion into orthopaedic literature or guidelines or through the study's average yearly citation rate.

RESULTS

One hundred sixty RCTs were included in the final analysis. A multivariate logistic regression identified large study sample size as the only predictor of an RCT being used in clinical guidance texts. Large sample size and multicenter RCTs were predictors of high yearly citation rates. The level of pragmatism in study design did not predict scholarly influence.

CONCLUSIONS

Pragmatic design is not independently associated with increased scholarly influence; however, large study sample size was the most important study characteristic affecting scholarly influence.

Examining Endpoint Concordance in Clinical Trials and Real-World Clinical Practice to Advance Real-World Evidence Utilization

Real-world evidence (RWE) is increasingly contributing to more informed decisions regarding the optimal access to and use of therapeutics to improve patient outcomes. However, in many cases, a disconnect between evidence derived from clinical trials and the RWE that follows market approval impedes the potential value and widespread adoption of RWE to optimize patient care. Collaborators with the Learning Ecosystems Accelerator for Patient-centered, Sustainable innovation (LEAPS), a major project of the Tufts Medical Center [formally Massachusetts Institute of Technology (MIT)] NEW Drug Development ParadIGmS (NEWDIGS) initiative, propose assessing the relationship between efficacy endpoints used in randomized controlled trials (RCTs) and effectiveness measures that inform treatment decisions within real-world clinical settings as one way to bridge this divide and further leverage RWE to improve care and patient outcomes. This commentary outlines elements of an endpoint concordance study using Rheumatoid Arthritis as a case study. The authors describe the ways in which better understanding of the relationship between effectiveness and RCT endpoints could improve the confidence in and adoption of RWE by both contextualizing existing RWE as well as identifying ways in which to improve the value of RWE in improving care and outcomes.

Consumer perspectives on simplified, layered consent for a low risk, but complex pragmatic trial

BACKGROUND

For decades, the research community has called for participant information sheets/consent forms (PICFs) to be improved. Recommendations include simplifying content, reducing length, presenting information in layers and using multimedia. However, there are relatively few studies that have evaluated health consumers' (patients/carers) perspectives on the type and organisation of information, and the level of detail to be included in a PICF to optimise an informed decision to enter a trial. We aimed to elicit consumers' views on a layered approach to consent that provides the key information for decision-making in a short PICF (layer 1) with additional optional information that is accessed separately (layer 2). We also elicited consumers' views on the optimal content and layout of the layered consent materials for a large and complex Bayesian adaptive platform trial (the SNAP trial).

METHODS

We conducted a qualitative multicentre study (4 focus groups and 2 semi-structured interviews) involving adolescent and adult survivors of Staphylococcus aureus bloodstream infection (22) and their carers (2). Interview transcripts were examined using inductive thematic analysis.

RESULTS

Consumers supported a layered approach to consent. The primary theme that emerged was the value of agency; the ability to exert some control over the amount of information read before the consent form is signed. Three other themes emerged; the need to prioritise participants' information needs; the importance of health literacy; the importance of information about a trial's benefits (over its risks) for decision-making and the interplay between the two.

CONCLUSIONS

Our findings suggest that consumers may challenge the one-size-fits-all approach currently applied to the development of PICFs in countries like Australia. Consumers supported a layered approach to consent that offers choice in the amount of information to be read before deciding whether to enter a trial. A 3-page PICF was considered sufficient for decision-making for the SNAP trial, provided that further information was available and accessible.

Deep Learning Artificial Intelligence to Predict the Need for Tracheostomy in Patients of Deep Neck Infection Based on Clinical and Computed Tomography Findings—Preliminary Data and a Pilot Study

Background: Deep neck infection (DNI) can lead to airway obstruction. Rather than intubation, some patients need tracheostomy to secure the airway. However, no study has used deep learning (DL) artificial intelligence (AI) to predict the need for tracheostomy in DNI patients. Thus, the purpose of this study was to develop a DL framework to predict the need for tracheostomy in DNI patients. Methods: 392 patients with DNI were enrolled in this study between August 2016 and April 2022; 80% of the patients (n = 317) were randomly assigned to a training group for model validation, and the remaining 20% (n = 75) were assigned to the test group to determine model accuracy. The k-nearest neighbor method was applied to analyze the clinical and computed tomography (CT) data of the patients. The predictions of the model with regard to the need for tracheostomy were compared with actual decisions made by clinical experts. Results: No significant differences were observed in clinical or CT parameters between the training group and test groups. The DL model yielded a prediction accuracy of 78.66% (59/75 cases). The sensitivity and specificity values were 62.50% and 80.60%, respectively. Conclusions: We demonstrated a DL framework to predict the need for tracheostomy in DNI patients based on clinical and CT data. The model has potential for clinical application; in particular, it may assist less experienced clinicians to determine whether tracheostomy is necessary in cases of DNI.

Effectiveness of Casirivimab-Imdevimab and Sotrovimab During a SARS-CoV-2 Delta Variant Surge: A Cohort Study and Randomized Comparative Effectiveness Trial

Importance

The effectiveness of monoclonal antibodies (mAbs), casirivimab-imdevimab and sotrovimab, is unknown in patients with mild to moderate COVID-19 caused by the SARS-CoV-2 Delta variant.

Objective

To evaluate the effectiveness of mAb against the Delta variant compared with no mAb treatment and to ascertain the comparative effectiveness of casirivimab-imdevimab and sotrovimab.

Design, Setting, and Participants

This study comprised 2 parallel studies: (1) a propensity score-matched cohort study of mAb treatment vs no mAb treatment and (2) a randomized comparative effectiveness trial of casirivimab-imdevimab and sotrovimab. The cohort consisted of patients who received mAb treatment at the University of Pittsburgh Medical Center outpatient infusion centers and emergency departments from July 14 to September 29, 2021. Participants were patients with a positive SARS-CoV-2 test result who were eligible to receive mAbs according to emergency use authorization criteria.

Exposure

For the trial, patients were randomized to either intravenous casirivimab-imdevimab or sotrovimab according to a system therapeutic interchange policy.

Main Outcomes and Measures

For the cohort study, risk ratio (RR) estimates for the primary outcome of hospitalization or death by 28 days were compared between mAb treatment and no mAb treatment using propensity score-matched models. For the comparative effectiveness trial, the primary outcome was hospital-free days (days alive and free of hospitalization) within 28 days after mAb treatment, where patients who died were assigned -1 day in a bayesian cumulative logistic model adjusted for treatment location, age, sex, and time. Inferiority was defined as a 99% posterior probability of an odds ratio (OR) less than 1. Equivalence was defined as a 95% posterior probability that the OR was within a given bound.

Results

A total of 3069 patients (1023 received mAb treatment: mean [SD] age, 53.2 [16.4] years; 569 women [56%]; 2046 had no mAb treatment: mean [SD] age, 52.8 [19.5] years; 1157 women [57%]) were included in the prospective cohort study, and 3558 patients (mean [SD] age, 54 [18] years; 1919 women [54%]) were included in the randomized comparative effectiveness trial. In propensity score-matched models, mAb treatment was associated with reduced risk of hospitalization or death (RR, 0.40; 95% CI, 0.28-0.57) compared with no treatment. Both casirivimab-imdevimab (RR, 0.31; 95% CI, 0.20-0.50) and sotrovimab (RR, 0.60; 95% CI, 0.37-1.00) were associated with reduced hospitalization or death compared with no mAb treatment. In the clinical trial, 2454 patients were randomized to receive casirivimab-imdevimab and 1104 patients were randomized to receive sotrovimab. The median (IQR) hospital-free days were 28 (28-28) for both mAb treatments, the 28-day mortality rate was less than 1% (n = 12) for casirivimab-imdevimab and less than 1% (n = 7) for sotrovimab, and the hospitalization rate by day 28 was 12% (n = 291) for casirivimab-imdevimab and 13% (n = 140) for sotrovimab. Compared with patients who received casirivimab-imdevimab, those who received sotrovimab had a median adjusted OR for hospital-free days of 0.88 (95% credible interval, 0.70-1.11). This OR yielded 86% probability of inferiority for sotrovimab vs casirivimab-imdevimab and 79% probability of equivalence.

Conclusions and Relevance

In this propensity score-matched cohort study and randomized comparative effectiveness trial, the effectiveness of casirivimab-imdevimab and sotrovimab against the Delta variant was similar, although the prespecified criteria for statistical inferiority or equivalence were not met. Both mAb treatments were associated with a reduced risk of hospitalization or death in nonhospitalized patients with mild to moderate COVID-19 caused by the Delta variant.

Trial Registration

ClinicalTrials.gov Identifier: NCT04790786.

The comparative effectiveness of COVID-19 monoclonal antibodies: A learning health system randomized clinical trial

Background

Monoclonal antibodies (mAb) that neutralize SARS-CoV-2 decrease hospitalization and death compared to placebo in patients with mild to moderate COVID-19; however, comparative effectiveness is unknown. We report the comparative effectiveness of bamlanivimab, bamlanivimab-etesevimab, and casirivimab-imdevimab.

Methods

A learning health system platform trial in a U.S. health system enrolled patients meeting mAb Emergency Use Authorization criteria. An electronic health record-embedded application linked local mAb inventory to patient encounters and provided random mAb allocation. Primary outcome was hospital-free days to day 28. Primary analysis was a Bayesian model adjusting for treatment location, age, sex, and time. Inferiority was defined as 99% posterior probability of an odds ratio < 1. Equivalence was defined as 95% posterior probability the odds ratio is within a given bound.

Findings

Between March 10 and June 25, 2021, 1935 patients received treatment. Median hospital-free days were 28 (IQR 28, 28) for each mAb. Mortality was 0.8% (1/128), 0.8% (7/885), and 0.7% (6/922) for bamlanivimab, bamlanivimab-etesevimab, and casirivimab-imdevimab, respectively. Relative to casirivimab-imdevimab (n = 922), median adjusted odds ratios were 0.58 (95% credible interval [CI] 0.30–1.16) and 0.94 (95% CI 0.72–1.24) for bamlanivimab (n = 128) and bamlanivimab-etesevimab (n = 885), respectively. These odds ratios yielded 91% and 94% probabilities of inferiority of bamlanivimab versus bamlanivimab-etesevimab and casirivimab-imdevimab, and an 86% probability of equivalence between bamlanivimab-etesevimab and casirivimab-imdevimab.

Interpretation

Among patients with mild to moderate COVID-19, bamlanivimab-etesevimab or casirivimab-imdevimab treatment resulted in 86% probability of equivalence. No treatment met prespecified criteria for statistical equivalence. Median hospital-free days to day 28 were 28 (IQR 28, 28) for each mAb.

Funding and registration

This work received no external funding. The U.S. government provided the reported mAb. This trial is registered at ClinicalTrials.gov, NCT04790786.

Deep learning model developed by multiparametric MRI in differential diagnosis of parotid gland tumors

PURPOSE

To create a new artificial intelligence approach based on deep learning (DL) from multiparametric MRI in the differential diagnosis of common parotid tumors.

METHODS

Parotid tumors were classified using the InceptionResNetV2 DL model and majority voting approach with MRI images of 123 patients. The study was conducted in three stages. At stage I, the classification of the control, pleomorphic adenoma, Warthin tumor and malignant tumor (MT) groups was examined, and two approaches in which MRI sequences were given in combined and non-combined forms were established. At stage II, the classification of the benign tumor, MT and control groups was made. At stage III, patients with a tumor in the parotid gland and those with a healthy parotid gland were classified.

RESULTS

A stage I, the accuracy value for classification in the non-combined and combined approaches was 86.43% and 92.86%, respectively. This value at stage II and stage III was found respectively as 92.14% and 99.29%.

CONCLUSIONS

The approach presented in this study classifies parotid tumors automatically and with high accuracy using DL models.

Combining factorial and multi-arm multi-stage platform designs to evaluate multiple interventions efficiently

BACKGROUND

Factorial-MAMS design platform designs have many advantages, but the practical advantages and disadvantages of combining the two designs have not been explored.

METHODS

We propose practical methods for a combined design within the platform trial paradigm where some interventions are not expected to interact and could be given together.

RESULTS

We describe the combined design and suggest diagrams that can be used to represent it. Many properties are common both to standard factorial designs, including the need to consider interactions between interventions and the impact of intervention efficacy on power of other comparisons, and to standard multi-arm multi-stage designs, including the need to pre-specify procedures for starting and stopping intervention comparisons. We also identify some specific features of the factorial-MAMS design: timing of interim and final analyses should be determined by calendar time or total observed events; some non-factorial modifications may be useful; eligibility criteria should be broad enough to include any patient eligible for any part of the randomisation; stratified randomisation may conveniently be performed sequentially; and analysis requires special care to use only concurrent controls.

CONCLUSION

A combined factorial-MAMS design can combine the efficiencies of factorial trials and multi-arm multi-stage platform trials. It allows us to address multiple research questions under one protocol and to test multiple new treatment options, which is particularly important when facing a new emergent infection such as COVID-19.

Cannabis, cannabinoids, and cannabis-based medicines: future research directions for analgesia

The use of cannabis spans thousands of years and encompasses almost all dimensions of the human experience, including consumption for recreational, religious, social, and medicinal purposes. Its use in the management of pain has been anecdotally described for millennia. However, an evidence base has only developed over the last 100 years, with an explosion in research occurring in the last 20-30 years, as more states in the USA as well as countries worldwide have legalized and encouraged its use in pain management. Pain remains one of the most common reasons for individuals deciding to use cannabis medicinally. However, cannabis remains illegal at the federal level in the USA and in most countries of the world, making it difficult to advance quality research on its efficacy for pain treatment. Nonetheless, new products derived both from the cannabis plant and the chemistry laboratory are being developed for use as analgesics. This review examines the current landscape of cannabinoids research and future research directions in the management of pain.

The principles of whole-hospital predictive analytics monitoring for clinical medicine originated in the neonatal ICU

In 2011, a multicenter group spearheaded at the University of Virginia demonstrated reduced mortality from real-time continuous cardiorespiratory monitoring in the neonatal ICU using what we now call Artificial Intelligence, Big Data, and Machine Learning. The large, randomized heart rate characteristics trial made real, for the first time that we know of, the promise that early detection of illness would allow earlier and more effective intervention and improved patient outcomes. Currently, though, we hear as much of failures as we do of successes in the rapidly growing field of predictive analytics monitoring that has followed. This Perspective aims to describe the principles of how we developed heart rate characteristics monitoring for neonatal sepsis and then applied them throughout adult ICU and hospital medicine. It primarily reflects the work since the 1990s of the University of Virginia group: the theme is that sudden and catastrophic deteriorations can be preceded by subclinical but measurable physiological changes apparent in the continuous cardiorespiratory monitoring and electronic health record.

Learning in times of stress: Lessons from COVID-19 that will last throughout this century

Systems tend toward inertia until an external pressure pushes them toward change; thus, a situation of crisis such as the COVID-19 pandemic represents an opportunity for technological innovation. The prevailing need for treatments and vaccines has impelled innovation in the world of randomized clinical trials (RCT), resorting to ideas that had been floating around for a while. Is this merely a circumstantial phenomenon or are new methods here to stay?

OUP accepted manuscript

The increasing volume and richness of healthcare data collected during routine clinical practice have not yet translated into significant numbers of actionable insights that have systematically improved patient outcomes. An evidence-practice gap continues to exist in healthcare. We contest that this gap can be reduced by assessing the use of nudge theory as part of clinical decision support systems (CDSS). Deploying nudges to modify clinician behaviour and improve adherence to guideline-directed therapy represents an underused tool in bridging the evidence-practice gap. In conjunction with electronic health records (EHRs) and newer devices including artificial intelligence algorithms that are increasingly integrated within learning health systems, nudges such as CDSS alerts should be iteratively tested for all stakeholders involved in health decision-making: clinicians, researchers, and patients alike. Not only could they improve the implementation of known evidence, but the true value of nudging could lie in areas where traditional randomized controlled trials are lacking, and where clinical equipoise and variation dominate. The opportunity to test CDSS nudge alerts and their ability to standardize behaviour in the face of uncertainty may generate novel insights and improve patient outcomes in areas of clinical practice currently without a robust evidence base.

COVID-19 and the generation of novel scientific knowledge: Research questions and study designs

RATIONALE, AIMS, AND OBJECTIVES

One of the sectors challenged by the COVID-19 pandemic is medical research. COVID-19 originates from a novel coronavirus (SARS-CoV-2) and the scientific community is faced with the daunting task of creating a novel model for this pandemic or, in other words, creating novel science. This paper is the first part of a series of two papers that explore the intricate relationship between the different challenges that have hindered biomedical research and the generation of scientific knowledge during the COVID-19 pandemic.

METHODS

During the early stages of the pandemic, research conducted on hydroxychloroquine (HCQ) was chaotic and sparked several heated debates with respect to the scientific methods used and the quality of knowledge generated. Research on HCQ is used as a case study in both papers. The authors explored biomedical databases, peer-reviewed journals, pre-print servers, and media articles to identify relevant literature on HCQ and COVID-19, and examined philosophical perspectives on medical research in the context of this pandemic and previous global health challenges.

RESULTS

This paper demonstrates that a lack of prioritization among research questions and therapeutics was responsible for the duplication of clinical trials and the dispersion of precious resources. Study designs, aimed at minimising biases and increasing objectivity, were, instead, the subject of fruitless oppositions. The duplication of research works, combined with poor-quality research, has greatly contributed to slowing down the creation of novel scientific knowledge.

CONCLUSIONS

The COVID-19 pandemic presented challenges in terms of (1) finding and prioritising relevant research questions and (2) choosing study designs that are appropriate for a time of emergency.

Early identification of patients with acute gastrointestinal bleeding using natural language processing and decision rules

BACKGROUND AND AIM

Guidelines recommend risk stratification scores in patients presenting with gastrointestinal bleeding (GIB), but such scores are uncommonly employed in practice. Automation and deployment of risk stratification scores in real time within electronic health records (EHRs) would overcome a major impediment. This requires an automated mechanism to accurately identify ("phenotype") patients with GIB at the time of presentation. The goal is to identify patients with acute GIB by developing and evaluating EHR-based phenotyping algorithms for emergency department (ED) patients.

METHODS

We specified criteria using structured data elements to create rules for identifying patients and also developed multiple natural language processing (NLP)-based approaches for automated phenotyping of patients, tested them with tenfold cross-validation for 10 iterations (n = 7144) and external validation (n = 2988) and compared them with a standard method to identify patient conditions, the Systematized Nomenclature of Medicine. The gold standard for GIB diagnosis was the independent dual manual review of medical records. The primary outcome was the positive predictive value.

RESULTS

A decision rule using GIB-specific terms from ED triage and ED review-of-systems assessment performed better than the Systematized Nomenclature of Medicine on internal validation and external validation (positive predictive value = 85% confidence interval:83%-87% vs 69% confidence interval:66%-72%; P < 0.001). The syntax-based NLP algorithm and Bidirectional Encoder Representation from Transformers neural network-based NLP algorithm had similar performance to the structured-data fields decision rule.

CONCLUSIONS

An automated decision rule employing GIB-specific triage and review-of-systems terms can be used to trigger EHR-based deployment of risk stratification models to guide clinical decision making in real time for patients with acute GIB presenting to the ED.

Strengthening the Learning Health System in Cardiovascular Disease Prevention: Time to Leverage Big Data and Digital Solutions

PURPOSE OF REVIEW

The past few decades have seen significant technologic innovation for the treatment and diagnosis of cardiovascular diseases. The subsequent growing complexity of modern medicine, however, is causing fundamental challenges in our healthcare system primarily in the spheres of patient involvement, data generation, and timely clinical implementation. The Institute of Medicine advocated for a learning health system (LHS) in which knowledge generation and patient care are inherently symbiotic. The purpose of this paper is to review how the advances in technology and big data have been used to further patient care and data generation and what future steps will need to occur to develop a LHS in cardiovascular disease.

RECENT FINDINGS

Patient-centered care has progressed from technologic advances yielding resources like decision aids. LHS can also incorporate patient preferences by increasing and standardizing patient-reported information collection. Additionally, data generation can be optimized using big data analytics by developing large interoperable datasets from multiple sources to allow for real-time data feedback. Developing a LHS will require innovative technologic solutions with a patient-centered lens to facilitate symbiosis in data generation and clinical practice.

The concept of justifiable healthcare and how big data can help us to achieve it

Over the last decades, the face of health care has changed dramatically, with big improvements in what is technically feasible. However, there are indicators that the current approach to evaluating evidence in health care is not holistic and hence in the long run, health care will not be sustainable. New conceptual and normative frameworks for the evaluation of health care need to be developed and investigated. The current paper presents a novel framework of justifiable health care and explores how the use of artificial intelligence and big data can contribute to achieving the goals of this framework.

Implementation of the Randomized Embedded Multifactorial Adaptive Platform for COVID-19 (REMAP-COVID) trial in a US health system-lessons learned and recommendations

Background

The Randomized Embedded Multifactorial Adaptive Platform for COVID-19 (REMAP-COVID) trial is a global adaptive platform trial of hospitalized patients with COVID-19. We describe implementation at the first US site, the UPMC health system, and offer recommendations for implementation at other sites.

Methods

To implement REMAP-COVID, we focused on six major areas: engaging leadership, trial embedment, remote consent and enrollment, regulatory compliance, modification of traditional trial management procedures, and alignment with other COVID-19 studies.

Results

We recommend aligning institutional and trial goals and sharing a vision of REMAP-COVID implementation as groundwork for learning health system development. Embedment of trial procedures into routine care processes, existing institutional structures, and the electronic health record promotes efficiency and integration of clinical care and clinical research. Remote consent and enrollment can be facilitated by engaging bedside providers and leveraging institutional videoconferencing tools. Coordination with the central institutional review board will expedite the approval process. Protocol adherence, adverse event monitoring, and data collection and export can be facilitated by building electronic health record processes, though implementation can start using traditional clinical trial tools. Lastly, establishment of a centralized institutional process optimizes coordination of COVID-19 studies.

Conclusions

Implementation of the REMAP-COVID trial within a large US healthcare system is feasible and facilitated by multidisciplinary collaboration. This investment establishes important groundwork for future learning health system endeavors.

Trial registration

NCT02735707. Registered on 13 April 2016.

On using electronic health records to improve optimal treatment rules in randomized trials

Individualized treatment rules (ITRs) tailor medical treatments according to patient-specific characteristics in order to optimize patient outcomes. Data from randomized controlled trials (RCTs) are used to infer valid ITRs using statistical and machine learning methods. However, RCTs are usually conducted under specific inclusion/exclusion criteria, thus limiting their generalizability to a broader patient population in real-world practice settings. Because electronic health records (EHRs) document treatment prescriptions in the real world, transferring information in EHRs to RCTs, if done appropriately, could potentially improve the performance of ITRs, in terms of precision and generalizability. In this work, we propose a new domain adaptation method to learn ITRs by incorporating information from EHRs. Unless we assume that there is no unmeasured confounding in EHRs, we cannot directly learn the optimal ITR from the combined EHR and RCT data. Instead, we first pretrain "super" features from EHRs that summarize physician treatment decisions and patient observed benefits in the real world, as these are likely to be informative of the optimal ITRs. We then augment the feature space of the RCT and learn the optimal ITRs by stratifying by super features using subjects enrolled in RCT. We adopt Q-learning and a modified matched-learning algorithm for estimation. We present heuristic justification of our method and conduct simulation studies to demonstrate the performance of super features. Finally, we apply our method to transfer information learned from EHRs of patients with type 2 diabetes to learn individualized insulin therapies from RCT data.

Interventions and Operations 5 Years After Bariatric Surgery in a Cohort From the US National Patient-Centered Clinical Research Network Bariatric Study

Importance

Additional data comparing longer-term problems associated with various bariatric surgical procedures are needed for shared decision-making.

Objective

To compare the risks of intervention, operation, endoscopy, hospitalization, and mortality up to 5 years after 2 bariatric surgical procedures.

Design, Setting, and Participants

Adults who underwent Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG) between January 1, 2005, and September 30, 2015, within the National Patient-Centered Clinical Research Network. Data from 33 560 adults at 10 centers within 4 clinical data research networks were included in this cohort study. Information was extracted from electronic health records using a common data model and linked to insurance claims and mortality indices. Analyses were conducted from January 2018 through October 2019.

Exposures

Bariatric surgical procedures.

Main Outcomes and Measures

The primary outcome was time until operation or intervention. Secondary outcomes included endoscopy, hospitalization, and mortality rates.

Results

Of 33 560 adults, 18 056 (54%) underwent RYGB, and 15 504 (46%) underwent SG. The median (interquartile range) follow-up for operation or intervention was 3.4 (1.6-5.0) years for RYGB and 2.2 (0.9-3.6) years for SG. The overall mean (SD) patient age was 45.0 (11.5) years, and the overall mean (SD) patient body mass index was 49.1 (7.9). The cohort was composed predominantly of women (80%) and white individuals (66%), with 26% of Hispanic ethnicity. Operation or intervention was less likely for SG than for RYGB (hazard ratio, 0.72; 95% CI, 0.65-0.79; P < .001). The estimated, adjusted cumulative incidence rates of operation or intervention at 5 years were 8.94% (95% CI, 8.23%-9.65%) for SG and 12.27% (95% CI, 11.49%-13.05%) for RYGB. Hospitalization was less likely for SG than for RYGB (hazard ratio, 0.82; 95% CI, 0.78-0.87; P < .001), and the 5-year adjusted cumulative incidence rates were 32.79% (95% CI, 31.62%-33.94%) for SG and 38.33% (95% CI, 37.17%-39.46%) for RYGB. Endoscopy was less likely for SG than for RYGB (hazard ratio, 0.47; 95% CI, 0.43-0.52; P < .001), and the adjusted cumulative incidence rates at 5 years were 7.80% (95% CI, 7.15%-8.43%) for SG and 15.83% (95% CI, 14.94%-16.71%) for RYGB. There were no differences in all-cause mortality between SG and RYGB.

Conclusions and Relevance

Interventions, operations, and hospitalizations were relatively common after bariatric surgical procedures and were more often associated with RYGB than SG.

Trial Registration

ClinicalTrials.gov identifier: NCT02741674.

A platform in the use of medicines to treat chronic hepatitis C (PLATINUM C): protocol for a prospective treatment registry of real-world outcomes for hepatitis C

Background

Safe, highly curative, short course, direct acting antiviral (DAA) therapies are now available to treat chronic hepatitis C. DAA therapy is freely available to all adults chronically infected with the hepatitis C virus (HCV) in Australia. If left untreated, hepatitis C may lead to progressive hepatic fibrosis, cirrhosis and hepatocellular carcinoma. Australia is committed to eliminating hepatitis as a public health threat by 2030 set by the World Health Organization. However, since the introduction of funded DAA treatment, uptake has been suboptimal. Australia needs improved strategies for testing, treatment uptake and treatment completion to address the persisting hepatitis C public health problem. PLATINUM C is a HCV treatment registry and research platform for assessing the comparative effectiveness of alternative interventions for achieving virological cure.

Methods

PLATINUM C will prospectively enrol people with active HCV infection confirmed by recent detection of HCV ribonucleic acid (RNA) in blood. Those enrolled will agree to allow standardised collection of demographic, lifestyle, treatment, virological outcome and other relevant clinical data to better inform the future management of HCV infection. The primary outcome is virological cure evidenced by sustained virological response (SVR), which is defined as a negative HCV PCR result 6 to 18 months after initial prescription of DAA therapy and no less than 12 weeks after the completion of treatment. Study participants will be invited to opt-in to medication adherence monitoring and quality of life assessments using validated self-reported instruments (EQ-5D-5L).

Discussion

PLATINUM C is a treatment registry and platform for nesting pragmatic trials. Data collected will inform the design, development and implementation of pragmatic trials. The digital infrastructure, study procedures and governing systems established by the registry will allow PLATINUM C to support a wider research platform in the management of hepatitis C in primary care.

Trial registration

The trial is registered with the Australia and New Zealand Clinical Trials Register (ACTRN12619000023156). Date of registration: 10/01/2019.

Supplementary Information

Supplementary information accompanies this paper at 10.1186/s12879-020-05531-4.

Effect of Hydrocortisone on Mortality and Organ Support in Patients With Severe COVID-19: The REMAP-CAP COVID-19 Corticosteroid Domain Randomized Clinical Trial

IMPORTANCE

Evidence regarding corticosteroid use for severe coronavirus disease 2019 (COVID-19) is limited.

OBJECTIVE

To determine whether hydrocortisone improves outcome for patients with severe COVID-19.

DESIGN, SETTING, AND PARTICIPANTS

An ongoing adaptive platform trial testing multiple interventions within multiple therapeutic domains, for example, antiviral agents, corticosteroids, or immunoglobulin. Between March 9 and June 17, 2020, 614 adult patients with suspected or confirmed COVID-19 were enrolled and randomized within at least 1 domain following admission to an intensive care unit (ICU) for respiratory or cardiovascular organ support at 121 sites in 8 countries. Of these, 403 were randomized to open-label interventions within the corticosteroid domain. The domain was halted after results from another trial were released. Follow-up ended August 12, 2020.

INTERVENTIONS

The corticosteroid domain randomized participants to a fixed 7-day course of intravenous hydrocortisone (50 mg or 100 mg every 6 hours) (n = 143), a shock-dependent course (50 mg every 6 hours when shock was clinically evident) (n = 152), or no hydrocortisone (n = 108).

MAIN OUTCOMES AND MEASURES

The primary end point was organ support-free days (days alive and free of ICU-based respiratory or cardiovascular support) within 21 days, where patients who died were assigned -1 day. The primary analysis was a bayesian cumulative logistic model that included all patients enrolled with severe COVID-19, adjusting for age, sex, site, region, time, assignment to interventions within other domains, and domain and intervention eligibility. Superiority was defined as the posterior probability of an odds ratio greater than 1 (threshold for trial conclusion of superiority >99%).

RESULTS

After excluding 19 participants who withdrew consent, there were 384 patients (mean age, 60 years; 29% female) randomized to the fixed-dose (n = 137), shock-dependent (n = 146), and no (n = 101) hydrocortisone groups; 379 (99%) completed the study and were included in the analysis. The mean age for the 3 groups ranged between 59.5 and 60.4 years; most patients were male (range, 70.6%-71.5%); mean body mass index ranged between 29.7 and 30.9; and patients receiving mechanical ventilation ranged between 50.0% and 63.5%. For the fixed-dose, shock-dependent, and no hydrocortisone groups, respectively, the median organ support-free days were 0 (IQR, -1 to 15), 0 (IQR, -1 to 13), and 0 (-1 to 11) days (composed of 30%, 26%, and 33% mortality rates and 11.5, 9.5, and 6 median organ support-free days among survivors). The median adjusted odds ratio and bayesian probability of superiority were 1.43 (95% credible interval, 0.91-2.27) and 93% for fixed-dose hydrocortisone, respectively, and were 1.22 (95% credible interval, 0.76-1.94) and 80% for shock-dependent hydrocortisone compared with no hydrocortisone. Serious adverse events were reported in 4 (3%), 5 (3%), and 1 (1%) patients in the fixed-dose, shock-dependent, and no hydrocortisone groups, respectively.

CONCLUSIONS AND RELEVANCE

Among patients with severe COVID-19, treatment with a 7-day fixed-dose course of hydrocortisone or shock-dependent dosing of hydrocortisone, compared with no hydrocortisone, resulted in 93% and 80% probabilities of superiority with regard to the odds of improvement in organ support-free days within 21 days. However, the trial was stopped early and no treatment strategy met prespecified criteria for statistical superiority, precluding definitive conclusions.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT02735707.

Effect of Hydrocortisone on Mortality and Organ Support in Patients With Severe COVID-19: The REMAP-CAP COVID-19 Corticosteroid Domain Randomized Clinical Trial

IMPORTANCE

Evidence regarding corticosteroid use for severe coronavirus disease 2019 (COVID-19) is limited.

OBJECTIVE

To determine whether hydrocortisone improves outcome for patients with severe COVID-19.

DESIGN, SETTING, AND PARTICIPANTS

An ongoing adaptive platform trial testing multiple interventions within multiple therapeutic domains, for example, antiviral agents, corticosteroids, or immunoglobulin. Between March 9 and June 17, 2020, 614 adult patients with suspected or confirmed COVID-19 were enrolled and randomized within at least 1 domain following admission to an intensive care unit (ICU) for respiratory or cardiovascular organ support at 121 sites in 8 countries. Of these, 403 were randomized to open-label interventions within the corticosteroid domain. The domain was halted after results from another trial were released. Follow-up ended August 12, 2020.

INTERVENTIONS

The corticosteroid domain randomized participants to a fixed 7-day course of intravenous hydrocortisone (50 mg or 100 mg every 6 hours) (n = 143), a shock-dependent course (50 mg every 6 hours when shock was clinically evident) (n = 152), or no hydrocortisone (n = 108).

MAIN OUTCOMES AND MEASURES

The primary end point was organ support-free days (days alive and free of ICU-based respiratory or cardiovascular support) within 21 days, where patients who died were assigned -1 day. The primary analysis was a bayesian cumulative logistic model that included all patients enrolled with severe COVID-19, adjusting for age, sex, site, region, time, assignment to interventions within other domains, and domain and intervention eligibility. Superiority was defined as the posterior probability of an odds ratio greater than 1 (threshold for trial conclusion of superiority >99%).

RESULTS

After excluding 19 participants who withdrew consent, there were 384 patients (mean age, 60 years; 29% female) randomized to the fixed-dose (n = 137), shock-dependent (n = 146), and no (n = 101) hydrocortisone groups; 379 (99%) completed the study and were included in the analysis. The mean age for the 3 groups ranged between 59.5 and 60.4 years; most patients were male (range, 70.6%-71.5%); mean body mass index ranged between 29.7 and 30.9; and patients receiving mechanical ventilation ranged between 50.0% and 63.5%. For the fixed-dose, shock-dependent, and no hydrocortisone groups, respectively, the median organ support-free days were 0 (IQR, -1 to 15), 0 (IQR, -1 to 13), and 0 (-1 to 11) days (composed of 30%, 26%, and 33% mortality rates and 11.5, 9.5, and 6 median organ support-free days among survivors). The median adjusted odds ratio and bayesian probability of superiority were 1.43 (95% credible interval, 0.91-2.27) and 93% for fixed-dose hydrocortisone, respectively, and were 1.22 (95% credible interval, 0.76-1.94) and 80% for shock-dependent hydrocortisone compared with no hydrocortisone. Serious adverse events were reported in 4 (3%), 5 (3%), and 1 (1%) patients in the fixed-dose, shock-dependent, and no hydrocortisone groups, respectively.

CONCLUSIONS AND RELEVANCE

Among patients with severe COVID-19, treatment with a 7-day fixed-dose course of hydrocortisone or shock-dependent dosing of hydrocortisone, compared with no hydrocortisone, resulted in 93% and 80% probabilities of superiority with regard to the odds of improvement in organ support-free days within 21 days. However, the trial was stopped early and no treatment strategy met prespecified criteria for statistical superiority, precluding definitive conclusions.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT02735707.

Strategies to Promote ResiliencY (SPRY): a randomised embedded multifactorial adaptative platform (REMAP) clinical trial protocol to study interventions to improve recovery after surgery in high-risk patients

INTRODUCTION

As the population ages, there is interest in strategies to promote resiliency, especially for frail patients at risk of its complications. The physiological stress of surgery in high-risk individuals has been proposed both as an important cause of accelerated age-related decline in health and as a model testing the effectiveness of strategies to improve resiliency to age-related health decline. We describe a randomised, embedded, multifactorial, adaptative platform (REMAP) trial to investigate multiple perioperative interventions, the first of which is metformin and selected for its anti-inflammatory and anti-ageing properties beyond its traditional blood glucose control features.

METHODS AND ANALYSIS

Within a multihospital, single healthcare system, the Core Protocol for Strategies to Promote ResiliencY (SPRY) will be embedded within both the electronic health record (EHR) and the healthcare culture generating a continuously self-learning healthcare system. Embedding reduces the administrative burden of a traditional trial while accessing and rapidly analysing routine patient care EHR data. SPRY-Metformin is a placebo-controlled trial and is the first SPRY domain evaluating the effectiveness of three metformin dosages across three preoperative durations within a heterogeneous set of major surgical procedures. The primary outcome is 90-day hospital-free days. Bayesian posterior probabilities guide interim decision-making with predefined rules to determine stopping for futility or superior dosing selection. Using response adaptative randomisation, a maximum of 2500 patients allows 77%-92% power, detecting >15% primary outcome improvement. Secondary outcomes include mortality, readmission and postoperative complications. A subset of patients will be selected for substudies evaluating the microbiome, cognition, postoperative delirium and strength.

ETHICS AND DISSEMINATION

The Core Protocol of SPRY REMAP and associated SPRY-Metformin Domain-Specific Appendix have been ethically approved by the Institutional Review Board and are publicly registered. Results will be publicly available to healthcare providers, patients and trial participants following achieving predetermined platform conclusions.

TRIAL REGISTRATION NUMBER

NCT03861767.

Data-driven ICU management: Using Big Data and algorithms to improve outcomes

The digitalization of the Intensive Care Unit (ICU) led to an increasing amount of clinical data being collected at the bedside. The term "Big Data" can be used to refer to the analysis of these datasets that collect enormous amount of data of different origin and format. Complexity and variety define the value of Big Data. In fact, the retrospective analysis of these datasets allows to generate new knowledge, with consequent potential improvements in the clinical practice. Despite the promising start of Big Data analysis in medical research, which has seen a rising number of peer-reviewed articles, very limited applications have been used in ICU clinical practice. A close future effort should be done to validate the knowledge extracted from clinical Big Data and implement it in the clinic. In this article, we provide an introduction to Big Data in the ICU, from data collection and data analysis, to the main successful examples of prognostic, predictive and classification models based on ICU data. In addition, we focus on the main challenges that these models face to reach the bedside and effectively improve ICU care.

Medication Management in Frail Older People: Consensus Principles for Clinical Practice, Research, and Education

Frailty is a geriatric condition associated with increased vulnerability to adverse drug events and medication-related harm. Existing clinical practice guidelines rarely provide medication management recommendations specific to frail older people. This report presents international consensus principles, generated by the Optimizing Geriatric Pharmacotherapy through Pharmacoepidemiology Network, related to medication management in frail older people. This consensus comprises 7 principles for clinical practice, 6 principles for research, and 4 principles for education. Principles for clinical practice include (1) perform medication reconciliation and maintain an up-to-date medication list; (2) assess and plan based on individual's capacity to self-manage medications; (3) ensure appropriate prescribing and deprescribing; (4) simplify medication regimens when appropriate to reduce unnecessary burden; (5) be alert to the contribution of medications to geriatric syndromes; (6) regularly review medication regimens to align with changing goals of care; and (7) facilitate multidisciplinary communication among patients, caregivers, and healthcare teams. Principles for research include (1) include frail older people in randomized controlled trials; (2) consider frailty status as an effect modifier; (3) ensure collection and reporting of outcome measures important in frailty; (4) assess impact of frailty on pharmacokinetics and pharmacodynamics; (5) encourage frailty research in under-researched settings; and (6) utilize routinely collected linked health data. Principles for education include (1) provide undergraduate and postgraduate education on frailty; (2) minimize low-value care related to medication management; (3) improve health and medication literacy; and (4) incorporate evidence in relation to frailty into clinical practice guidelines. These principles for clinical practice, research and education highlight different considerations for optimizing medication management in frail older people. These principles can be used in conjunction with existing best practice guidelines to help achieve optimal health outcomes for this vulnerable population. Implementation of the principles will require multidisciplinary collaboration between healthcare professionals, researchers, educators, organizational leaders, and policymakers.

The REMAP-CAP (Randomized Embedded Multifactorial Adaptive Platform for Community-acquired Pneumonia) Study. Rationale and Design

There is broad interest in improved methods to generate robust evidence regarding best practice, especially in settings where patient conditions are heterogenous and require multiple concomitant therapies. Here, we present the rationale and design of a large, international trial that combines features of adaptive platform trials with pragmatic point-of-care trials to determine best treatment strategies for patients admitted to an intensive care unit with severe community-acquired pneumonia. The trial uses a novel design, entitled "a randomized embedded multifactorial adaptive platform." The design has five key features: 1) randomization, allowing robust causal inference; 2) embedding of study procedures into routine care processes, facilitating enrollment, trial efficiency, and generalizability; 3) a multifactorial statistical model comparing multiple interventions across multiple patient subgroups; 4) response-adaptive randomization with preferential assignment to those interventions that appear most favorable; and 5) a platform structured to permit continuous, potentially perpetual enrollment beyond the evaluation of the initial treatments. The trial randomizes patients to multiple interventions within four treatment domains: antibiotics, antiviral therapy for influenza, host immunomodulation with extended macrolide therapy, and alternative corticosteroid regimens, representing 240 treatment regimens. The trial generates estimates of superiority, inferiority, and equivalence between regimens on the primary outcome of 90-day mortality, stratified by presence or absence of concomitant shock and proven or suspected influenza infection. The trial will also compare ventilatory and oxygenation strategies, and has capacity to address additional questions rapidly during pandemic respiratory infections. As of January 2020, REMAP-CAP (Randomized Embedded Multifactorial Adaptive Platform for Community-acquired Pneumonia) was approved and enrolling patients in 52 intensive care units in 13 countries on 3 continents. In February, it transitioned into pandemic mode with several design adaptations for coronavirus disease 2019. Lessons learned from the design and conduct of this trial should aid in dissemination of similar platform initiatives in other disease areas.Clinical trial registered with www.clinicaltrials.gov (NCT02735707).

Neurocritical Care: Bench to Bedside (Eds. Claude Hemphill, Michael James) Integrating and Using Big Data in Neurocritical Care

The critical care environment drives huge volumes of data, and clinicians are tasked with quickly processing this data and responding to it urgently. The neurocritical care environment increasingly involves EEG, multimodal intracranial monitoring, and complex imaging which preclude comprehensive human synthesis, and requires new concepts to integrate data into clinical care. By definition, Big Data is data that cannot be handled using traditional infrastructures and is characterized by the volume, variety, velocity, and variability of the data being produced. Big Data in the neurocritical care unit requires rethinking of data storage infrastructures and the development of tools and analytics to drive advancements in the field. Preprocessing, feature extraction, statistical inference, and analytic tools are required in order to achieve the primary goals of Big Data for clinical use: description, prediction, and prescription. Barriers to its use at bedside include a lack of infrastructure development within the healthcare industry, lack of standardization of data inputs, and ultimately existential and scientific concerns about the outputs that result from the use of tools such as artificial intelligence. However, as implied by the fundamental theorem of biomedical informatics, physicians remain central to the development and utility of Big Data to improve patient care.

ICU management based on big data

PURPOSE OF REVIEW

The availability of large datasets and computational power has prompted a revolution in Intensive Care. Data represent a great opportunity for clinical practice, benchmarking, and research. Machine learning algorithms can help predict events in a way the human brain can simply not process. This possibility comes with benefits and risks for the clinician, as finding associations does not mean proving causality.

RECENT FINDINGS

Current applications of Data Science still focus on data documentation and visualization, and on basic rules to identify critical lab values. Recently, algorithms have been put in place for prediction of outcomes such as length of stay, mortality, and development of complications. These results have begun being implemented for more efficient allocation of resources and in benchmarking processes, to allow identification of successful practices and margins for improvement. In parallel, machine learning models are increasingly being applied in research to expand medical knowledge.

SUMMARY

Data have always been part of the work of intensivists, but the current availability has not been completely exploited. The intensive care community has to embrace and guide the data science revolution in order to decline it in favor of patients' care.

Identifying Clinical Research Priorities in Adult Pulmonary and Critical Care: NHLBI Working Group Report

Preventing, treating, and promoting recovery from critical illness due to pulmonary disease are foundational goals of the critical care community and the National Heart, Lung, and Blood Institute. Decades of clinical research in acute respiratory distress syndrome, acute respiratory failure, pneumonia, and sepsis have yielded improvements in supportive care, which have translated into improved patient outcomes. Novel therapeutics have largely failed to translate from promising pre-clinical findings into improved patient outcomes in late-phase clinical trials. Recent advances in personalized medicine, "big data", causal inference using observational data, novel clinical trial designs, pre-clinical disease modeling, and understanding recovery from acute illness promise to transform the methods of pulmonary and critical care clinical research. To assess the current state, research priorities, and future directions for adult pulmonary and critical care research, the NHLBI assembled a multidisciplinary working group of investigators. This working group identified recommendations for future research, including: (1) focusing on understanding the clinical, physiological, and biological underpinnings of heterogeneity in syndromes, diseases, and treatment-response with the goal of developing targeted, personalized interventions; (2) optimizing pre-clinical models by incorporating comorbidities, co-interventions, and organ support; (3) developing and applying novel clinical trial designs; and (4) advancing mechanistic understanding of injury and recovery in order to develop and test interventions targeted at achieving long-term improvements in the lives of patients and families. Specific areas of research are highlighted as especially promising for making advances in pneumonia, acute hypoxemic respiratory failure, and acute respiratory distress syndrome.

The Future of Data-Driven Wound Care

Care for patients with chronic wounds can be complex, and the chances of poor outcomes are high if wound care is not optimized through evidence-based protocols. Tracking and managing every variable and comorbidity in patients with wounds is difficult despite the increasing use of wound-specific electronic medical records. Harnessing the power of big data analytics to help nurses and physicians provide optimized care based on the care provided to millions of patients can result in better outcomes. Numerous applications of machine learning toward workflow improvements, inpatient monitoring, outpatient communication, and hospital operations can improve overall efficiency and efficacy of care delivery in and out of the hospital, while reducing adverse events and complications. This article provides an overview of the application of big data analytics and machine learning in health care, highlights important recent advances, and discusses how these technologies may revolutionize advanced wound care.

A framework for analysing learning health systems: Are we removing the most impactful barriers?

INTRODUCTION

Learning health systems (LHS) are one of the major computing advances in health care. However, no prior research has systematically analysed barriers and facilitators for LHS. This paper presents an investigation into the barriers, benefits, and facilitating factors for LHS in order to create a basis for their successful implementation and adoption.

METHODS

First, the ITPOSMO-BBF framework was developed based on the established ITPOSMO (information, technology, processes, objectives, staffing, management, and other factors) framework, extending it for analysing barriers, benefits, and facilitators. Second, the new framework was applied to LHS.

RESULTS

We found that LHS shares similar barriers and facilitators with electronic health records (EHR); in particular, most facilitator effort in implementing EHR and LHS goes towards barriers categorised as human factors, even though they were seen to carry fewer benefits. Barriers whose resolution would bring significant benefits in safety, quality, and health outcomes remain.LHS envisage constant generation of new clinical knowledge and practice based on the central role of collections of EHR. Once LHS are constructed and operational, they trigger new data streams into the EHR. So LHS and EHR have a symbiotic relationship. The implementation and adoption of EHRs have proved and continues to prove challenging, and there are many lessons for LHS arising from these challenges.

CONCLUSIONS

Successful adoption of LHS should take account of the framework proposed in this paper, especially with respect to its focus on removing barriers that have the most impact.

Randomized controlled trials reflected clinical practice when comparing the course of low back pain symptoms in similar populations

OBJECTIVE

This study compares participants in randomized controlled trials (RCTs) (the Minimal Invasive Treatment [MinT] trials) to participants in a related observational study with regard to their low back pain (LBP) symptom course.

STUDY DESIGN AND SETTING

Eligible patients were diagnosed with chronic LBP originating from the facet joints (N = 615) or sacroiliac (SI) joints (N = 533) and were treated with radiofrequency denervation and an exercise program. Randomized patients were compared to patients in the related observational study who fulfilled all RCT eligibility criteria (observational group 1) and to patients who did not fulfill at least one of the RCT eligibility criteria (observational group 2). Outcomes were pain intensity, treatment success, and functional status over a 3-month period. Longitudinal mixed-model analyses and linear regression models were applied to analyze the differences in outcomes between the RCT and observational study groups.

RESULTS

No differences in symptom course were found between patients in the RCTs and patients in observational group 1. Patients with facet joint pain in observational group 2 had overall less treatment success (odds ratios [OR], 0.67; 95% confidence interval [CI], 0.50-0.90), and less improvement in physical functioning (mean difference [MD], 5.82; 95% CI, 2.54-9.11) compared to the RCT patients. Patients with SI joint pain in observational group 2 had higher pain scores (MD, 0.40; 95% CI, 0.09-0.72), less treatment success (OR, 0.72; 95% CI, 0.54-0.96), and less improvement in physical functioning (MD, 7.16; 95% CI, 3.84-10.47) compared to the RCT patients.

CONCLUSION

This supports the generalizability of results from the MinT RCTs as this study suggests that these RCTs reflect clinical practice when comparing similar populations. To what extent this holds true for all RCTs in LBP should be further explored.

The future of probiotics in the preterm infant

Probiotic administration to preterm infants is not universal despite randomised trial data from >10,000 infants, significant observational data and multiple meta-analyses. Advocates point to reductions in necrotising enterocolitis and sepsis, 'sceptics' hold concerns over data quality/interpretation or risks. Issues revolve around different products, primary outcomes, uncertain dosing strategies and individual large 'negative' trials alongside probiotic associated sepsis and quality control concerns. We review concerns and how to move probiotic use forward. Surprisingly little is known about parental perspectives, vital to inform next steps. How to share information and decisions around probiotic use now, and how this impacts on future available strategies is discussed. We address placebo controlled trials and propose alternate designs, including head to head studies, using 'routine' data collection systems, opt out consents and 'learning technologies' embedded in health care systems. We also raise the importance of underpinning mechanistic work to inform future trials.