A clinically applicable approach to continuous prediction of future acute kidney injury.

The early prediction of deterioration could have an important role in supporting healthcare professionals, as an estimated 11% of deaths in hospital follow a failure to promptly recognize and treat deteriorating patients1. To achieve this goal requires predictions of patient risk that are continuously updated and accurate, and delivered at an individual level with sufficient context and enough time to act. Here we develop a deep learning approach for the continuous risk prediction of future deterioration in patients, building on recent work that models adverse events from electronic health records2-17 and using acute kidney injury-a common and potentially life-threatening condition18-as an exemplar. Our model was developed on a large, longitudinal dataset of electronic health records that cover diverse clinical environments, comprising 703,782 adult patients across 172 inpatient and 1,062 outpatient sites. Our model predicts 55.8% of all inpatient episodes of acute kidney injury, and 90.2% of all acute kidney injuries that required subsequent administration of dialysis, with a lead time of up to 48 h and a ratio of 2 false alerts for every true alert. In addition to predicting future acute kidney injury, our model provides confidence assessments and a list of the clinical features that are most salient to each prediction, alongside predicted future trajectories for clinically relevant blood tests9. Although the recognition and prompt treatment of acute kidney injury is known to be challenging, our approach may offer opportunities for identifying patients at risk within a time window that enables early treatment.

Digital health and artificial intelligence in kidney research: a report from the 2020 Kidney Disease Clinical Trialists (KDCT) meeting.

The exponential growth in digital technology coupled with the global coronavirus disease 2019 pandemic is driving a profound change in the delivery of medical care and research conduct. The growing availability of electronic monitoring, electronic health records, smartphones and other devices and access to ever greater computational power provides not only new opportunities, but also new challenges. Artificial intelligence (AI) exemplifies the potential of this digital revolution, which also includes other tools such as mobile health (mHealth) services and wearables. Despite digital technology becoming commonplace, its use in medicine and medical research is still in its infancy, with many clinicians and researchers having limited experience with such tools in their usual practice. This article, derived from the 'Digital Health and Artificial Intelligence' session of the Kidney Disease Clinical Trialists virtual workshop held in September 2020, aims to illustrate the breadth of applications to which digital tools and AI can be applied in clinical medicine and research. It highlights several innovative projects incorporating digital technology that range from streamlining medical care of those with acute kidney injury to the use of AI to navigate the vast genomic and proteomic data gathered in kidney disease. Important considerations relating to any new digital health project are presented, with a view to encouraging the further evolution and refinement of these new tools in a manner that fosters collaboration and the generation of robust evidence.

Glycine Amidinotransferase (GATM), Renal Fanconi Syndrome, and Kidney Failure.

Background For many patients with kidney failure, the cause and underlying defect remain unknown. Here, we describe a novel mechanism of a genetic order characterized by renal Fanconi syndrome and kidney failure.Methods We clinically and genetically characterized members of five families with autosomal dominant renal Fanconi syndrome and kidney failure. We performed genome-wide linkage analysis, sequencing, and expression studies in kidney biopsy specimens and renal cells along with knockout mouse studies and evaluations of mitochondrial morphology and function. Structural studies examined the effects of recognized mutations.Results The renal disease in these patients resulted from monoallelic mutations in the gene encoding glycine amidinotransferase (GATM), a renal proximal tubular enzyme in the creatine biosynthetic pathway that is otherwise associated with a recessive disorder of creatine deficiency. In silico analysis showed that the particular GATM mutations, identified in 28 members of the five families, create an additional interaction interface within the GATM protein and likely cause the linear aggregation of GATM observed in patient biopsy specimens and cultured proximal tubule cells. GATM aggregates-containing mitochondria were elongated and associated with increased ROS production, activation of the NLRP3 inflammasome, enhanced expression of the profibrotic cytokine IL-18, and increased cell death.Conclusions In this novel genetic disorder, fully penetrant heterozygous missense mutations in GATM trigger intramitochondrial fibrillary deposition of GATM and lead to elongated and abnormal mitochondria. We speculate that this renal proximal tubular mitochondrial pathology initiates a response from the inflammasome, with subsequent development of kidney fibrosis.

The definition of acute kidney injury and its use in practice.

Acute kidney injury (AKI) is a common syndrome that is independently associated with increased mortality. A standardized definition is important to facilitate clinical care and research. The definition of AKI has evolved rapidly since 2004, with the introduction of the Risk, Injury, Failure, Loss, and End-stage renal disease (RIFLE), AKI Network (AKIN), and Kidney Disease Improving Global Outcomes (KDIGO) classifications. RIFLE was modified for pediatric use (pRIFLE). They were developed using both evidence and consensus. Small rises in serum creatinine are independently associated with increased mortality, and hence are incorporated into the current definition of AKI. The recent definition from the international KDIGO guideline merged RIFLE and AKIN. Systematic review has found that these definitions do not differ significantly in their performance. Health-care staff caring for children or adults should use standard criteria for AKI, such as the pRIFLE or KDIGO definitions, respectively. These efforts to standardize AKI definition are a substantial advance, although areas of uncertainty remain. The new definitions have enabled the use of electronic alerts to warn clinicians of possible AKI. Novel biomarkers may further refine the definition of AKI, but their use will need to produce tangible improvements in outcomes and cost effectiveness. Further developments in AKI definitions should be informed by research into their practical application across health-care providers. This review will discuss the definition of AKI and its use in practice for clinicians and laboratory scientists.

The utility of ADAMTS13 in differentiating TTP from other acute thrombotic microangiopathies: results from the UK TTP Registry.

Thrombotic microangiopathies (TMAs) are frequently difficult to differentiate clinically, and measurement of ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13) remains vital in thrombotic thrombocytopenic purpura (TTP) diagnosis. We retrospectively reviewed cases referred for ADAMTS13 testing, using UK TTP Registry screening data. Of a total 810 cases, 350 were confirmed as TTP. The 460 non-TTP cases comprised secondary TMAs (24·57%) and haemolytic uraemic syndrome (HUS) (27·17% aHUS, 2·83% Shiga-like toxin-producing E. coli [STEC]-HUS); the remainder were TMAs with no clear association, not TMAs, or had no confirmed diagnosis. ADAMTS13 levels were significantly lower in TTP than STEC-HUS, aHUS and other TMAs. TTP patients had significantly lower platelet count (15 × 10(9) /l; range 0-96) than aHUS (57 × 10(9) /l; range 13-145, P < 0·0001) or STEC-HUS (35 × 10(9) /l; range 14-106, P < 0·0001); they also had lower creatinine levels (92 µmol/l; range 43-374) than aHUS (255 µmol/l; range 23-941, P < 0·0001) and STEC-HUS (324 µmol/l; range 117-639, P < 0·0001). However, 12/34 (35·3%) aHUS patients had a platelet count <30 × 10(9) /l and 26/150 (17·3%) of TTP patients had a platelet count >30 × 10(9) /l; 23/150 (15·3%) of TTP patients had a creatinine level >150 µmol/l. This study highlights the wide variety of TMA presentations, and confirms the utility of ADAMTS13 testing in TTP diagnosis.

Contrast-induced acute kidney injury following PCI.

BACKGROUND: Coronary revascularization using percutaneous coronary intervention (PCI) is one of the major treatments for patients with stable coronary artery disease, with approximately 1.5 million patients undergoing PCI in the United States and Europe every year. An important neglected complication of PCI is contrast-induced acute kidney injury (CI-AKI). DESIGN: In this article, we review the definition, pathogenesis and management of CI-AKI and highlight potential therapeutic options for preventing CI-AKI in post-PCI patients. RESULTS: CI-AKI is an important but underdiagnosed complication of PCI that is associated with increased in-hospital morbidity and mortality. Patients with pre-existing renal impairment and diabetes are particularly susceptible to this complication post-PCI. Optimization of the patients' circulating volume remains the mainstay for preventing CI-AKI, although the best strategy for achieving this is still controversial. CONCLUSION: Following PCI, CI-AKI is an overlooked complication which is associated with significant morbidity and mortality. In this article, we review the pathophysiology of CI-AKI in patients undergoing PCI and discuss the potential therapeutic options for preventing it.

The enzyme 4-hydroxy-2-oxoglutarate aldolase is deficient in primary hyperoxaluria type 3.

BACKGROUND: Mutations in the 4-hydroxy-2-oxoglutarate aldolase (HOGA1) gene have been recently identified in patients with atypical primary hyperoxaluria (PH). However, it was not clearly established whether these mutations caused disease via loss of function or activation of the gene product. METHODS: Whole-gene sequencing of HOGA1 was conducted in 28 unrelated patients with a high clinical suspicion of PH and in whom Types 1 and 2 had been excluded. RESULTS: Fifteen patients were homozygous or compound heterozygous for mutations in HOGA1. In total, seven different mutations were identified including three novel changes: a missense mutation, c.107C > T (p.Ala36Val), and two nonsense mutations c.117C > A (p.Tyr39X) and c.208C > T (p.Arg70X) as well as the previously documented c.860G > T (p.Gly297Val), c.907C > T (p.Arg303Cys) and in-frame c.944_946delAGG (p.Glu315del) mutations. The recurrent c.700 + 5G > T splice site mutation in intron 5 was most common with a frequency of 67%. Expression studies on hepatic messenger RNA demonstrated the pathogenicity of this mutation. CONCLUSIONS: The detection of a patient with two novel nonsense mutations within exon 1 of the gene, c.117C > A (p.Tyr39X) and c.208C > T (p.Arg70X), provides definitive proof that PH Type 3 is due to deficiency of the 4-hydroxy-2-oxoglutarate aldolase enzyme.

Comparison of renal replacement therapy and renal recovery before and during the COVID-19 pandemic: a single center observational study.

BACKGROUND: Our objective was to the describe indications, management, complications and outcomes of renal replacement therapy (RRT) in COVID-19 critically ill patients. To contextualize these findings, comparisons were made against 36 non-COVID-19 consecutive patients requiring RRT on ICU. METHODS: We conducted a retrospective single center observational cohort study of patients requiring acute RRT between 1st March and 30th June 2020. Comparison was made against those receiving RRT in the pre-COVID-19 period from January 2019 to February 2020. RESULTS: Of 154 COVID-19 patients, 47 (30.5%) received continuous venovenous hemofiltration (CVVHF), all of whom required mechanical ventilation and vasopressor support. The requirement for RRT was related to fluid balance rather than azotemia. Compared to 36 non-COVID-19 patients, those with COVID-19 were younger (P=0.016) with a lower serum creatinine on hospital admission (P=0.049), and lesser degrees of metabolic acidosis (P<0.001) and lactatemia (P<0.001) before initiation of RRT. In addition, the duration of RRT requirement was longer (P<0.001). Despite lower CVVHF exchange rates with higher serum creatinine levels following RRT initiation in the COVID-19 patients, metabolic abnormalities were corrected. Hospital mortality was 60% among COVID-19 patients requiring RRT, compared to 67% in non-COVID-19 patients (P=0.508), and renal recovery among survivors without pre-existing CKD was similar (P=0.231). CONCLUSIONS: The requirement for RRT in COVID-19 patients was primarily related to fluid balance. Using lower CVVHF exchange rates was effective to correct metabolic abnormalities. Renal recovery occurred in all but one patient by 60 days in the 40% of patients who survived.

Use of deep learning to develop continuous-risk models for adverse event prediction from electronic health records.

Early prediction of patient outcomes is important for targeting preventive care. This protocol describes a practical workflow for developing deep-learning risk models that can predict various clinical and operational outcomes from structured electronic health record (EHR) data. The protocol comprises five main stages: formal problem definition, data pre-processing, architecture selection, calibration and uncertainty, and generalizability evaluation. We have applied the workflow to four endpoints (acute kidney injury, mortality, length of stay and 30-day hospital readmission). The workflow can enable continuous (e.g., triggered every 6 h) and static (e.g., triggered at 24 h after admission) predictions. We also provide an open-source codebase that illustrates some key principles in EHR modeling. This protocol can be used by interdisciplinary teams with programming and clinical expertise to build deep-learning prediction models with alternate data sources and prediction tasks.

Effect of remote ischemic preconditioning on acute kidney injury in nondiabetic patients undergoing coronary artery bypass graft surgery: a secondary analysis of 2 small randomized trials.

BACKGROUND: Novel treatment strategies are required to reduce the development of acute kidney injury (AKI) in patients undergoing cardiac surgery. In this respect, remote ischemic preconditioning (RIPC), a phenomenon in which transient nonlethal ischemia applied to an organ or tissue protects another organ or tissue from subsequent lethal ischemic injury, is a potential renoprotective strategy. STUDY DESIGN: Secondary analysis of 2 randomized trials. SETTING & PARTICIPANTS: 78 consenting selected nondiabetic patients in a university teaching hospital undergoing elective coronary artery bypass graft (CABG) surgery recruited to 2 previously reported randomized studies. INTERVENTION: RIPC consisted of three 5-minute cycles of right forearm ischemia, induced by inflating a blood pressure cuff on the upper arm to 200 mm Hg, with an intervening 5 minutes of reperfusion, during which time the cuff was deflated. The control consisted of placing an uninflated cuff on the arm for 30 minutes. OUTCOMES: AKI measured using Acute Kidney Injury Network (AKIN) criteria, duration of hospital stay, in-hospital and 30-day mortality. RESULTS: Numbers of participants with AKI stages 1, 2, and 3 were 1 (3%), 3 (8%), and 0 in the intervention group compared with 10 (25%), 0, and 0 in the control group, respectively (P = 0.005). The decrease in AKI was independent of the effect of concomitant aortic valve replacement and cross-clamp times, which were distributed unevenly between the 2 groups. LIMITATIONS: Retrospective analysis of data. More patients in the RIPC group underwent concomitant aortic valve replacement with CABG; although we have corrected statistically for this imbalance, it remains an important confounding variable. CONCLUSIONS: RIPC induced using transient forearm ischemia decreased the incidence of AKI in nondiabetic patients undergoing elective CABG surgery in this retrospective analysis. A large prospective clinical trial is required to study this effect and clinical outcomes in patients undergoing cardiac surgery.

Evaluation of a digitally-enabled care pathway for acute kidney injury management in hospital emergency admissions.

We developed a digitally enabled care pathway for acute kidney injury (AKI) management incorporating a mobile detection application, specialist clinical response team and care protocol. Clinical outcome data were collected from adults with AKI on emergency admission before (May 2016 to January 2017) and after (May to September 2017) deployment at the intervention site and another not receiving the intervention. Changes in primary outcome (serum creatinine recovery to ≤120% baseline at hospital discharge) and secondary outcomes (30-day survival, renal replacement therapy, renal or intensive care unit (ICU) admission, worsening AKI stage and length of stay) were measured using interrupted time-series regression. Processes of care data (time to AKI recognition, time to treatment) were extracted from casenotes, and compared over two 9-month periods before and after implementation (January to September 2016 and 2017, respectively) using pre-post analysis. There was no step change in renal recovery or any of the secondary outcomes. Trends for creatinine recovery rates (estimated odds ratio (OR) = 1.04, 95% confidence interval (95% CI): 1.00-1.08, p = 0.038) and renal or ICU admission (OR = 0.95, 95% CI: 0.90-1.00, p = 0.044) improved significantly at the intervention site. However, difference-in-difference analyses between sites for creatinine recovery (estimated OR = 0.95, 95% CI: 0.90-1.00, p = 0.053) and renal or ICU admission (OR = 1.06, 95% CI: 0.98-1.16, p = 0.140) were not significant. Among process measures, time to AKI recognition and treatment of nephrotoxicity improved significantly (p < 0.001 and 0.047 respectively).

Building motivation to participate in a quality improvement collaborative in NHS hospital trusts in Southeast England: a qualitative participatory evaluation.

OBJECTIVES: This study explores the barriers and facilitators that impact on the motivation of practitioners to participate in a quality improvement collaborative. DESIGN: A qualitative and formative evaluation using a participatory approach, the researcher-in-residence model which embraces the concept of 'coproducing' knowledge between researchers and practitioners using a range of research methods such as participant observation, interviews and documentary analysis. The design, creation and application of newly generated evidence are facilitated by the researcher through negotiation and compromise with team members. PARTICIPANTS: Senior and middle managers, doctors and nurses. SETTING: Two hospitals in Southeast England participating in a Patient Safety Improvement Collaborative and the facilitator (host) of the collaborative, based in Central London. RESULTS: The evaluation has revealed facilitators and barriers to motivation categorised under two main themes: (1) inherent motivation and (2) factors that influence motivation, interorganisational and intraorganisational features as well as external factors. Facilitators included collaborative 'champions,' individuals who drove the quality improvement agenda at a local level, raising awareness and inspiring colleagues. The collaborative itself acted as a facilitator, promoting shared learning as well as building motivation for participation. A key barrier was the lack of board engagement in the participating National Health Service organisations which may have affected motivation among front-line staff. CONCLUSIONS: Collaboratives maybe an important way of engaging practitioners in quality improvement initiatives. This study highlights that despite a challenging healthcare environment in the UK, there remains motivation among individuals to participate in quality improvement programmes as they recognise that improvement approaches may facilitate positive change in local clinical processes and systems. Collaboratives can harness this individual motivation to facilitate spread and adoption of improvement methodology and build engagement across their membership.

Persistent severe polyuria after renal transplant.

Polydipsia and polyuria are common symptoms in patients with diabetes insipidus (DI), which can be due to inadequate vasopressin production (cranial DI) or vasopressin insensitivity (nephrogenic DI). Clinical diagnosis of the subtypes of DI can be tricky. We present a 44-year-old man with a strong family history of DI who had been diagnosed with autosomal dominant nephrogenic DI from infancy. At the age of 40, he had progressed to end-stage renal failure. When he experienced unresolving severe polyuria after renal transplant, further investigations revealed that he was misdiagnosed and that he had a novel mutation causing autosomal dominant cranial DI.

Acute kidney injury.

Acute kidney injury (AKI) - an abrupt deterioration in renal function - causes a rise in serum creatinine (SCr) or fall in urine output. It is common, occurring in up to 20% of hospital admissions. Importantly, even small rises in SCr are associated with increased risk of death and longer hospital stays. A 2009 National Confidential Enquiry into Patient Outcome and Death report found that a proportion of AKI in secondary care was avoidable. In addition, management of established AKI was 'good' less than half the time. In practice, AKI represents a heterogeneous group of conditions, encompassing impairments in both kidney structure and function. Delivering disease-specific treatment early in the course of AKI may improve outcomes. The provision of best-practice care for all will rely on a better understanding of risk, and frameworks of care that can be applied across a diverse patient group.

Effect of remote ischemic preconditioning on clinical outcomes in patients undergoing coronary artery bypass graft surgery (ERICCA): rationale and study design of a multi-centre randomized double-blinded controlled clinical trial.

BACKGROUND: Novel cardioprotective strategies are required to improve clinical outcomes in high risk patients undergoing coronary artery bypass graft (CABG) ± valve surgery. Remote ischemic preconditioning (RIC), in which brief episodes of non-lethal ischemia and reperfusion are applied to the arm or leg, has been demonstrated to reduce perioperative myocardial injury following CABG ± valve surgery. Whether RIC can improve clinical outcomes in this setting is unknown and is investigated in the effect of remote ischemic preconditioning on clinical outcomes (ERICCA) trial in patients undergoing CABG surgery. (ClinicalTrials.gov Identifier: NCT01247545). METHODS: The ERICCA trial is a multicentre randomized double-blinded controlled clinical trial which will recruit 1,610 high-risk patients (Additive Euroscore ≥ 5) undergoing CABG ± valve surgery using blood cardioplegia via 27 tertiary centres over 2 years. The primary combined endpoint will be cardiovascular death, non-fatal myocardial infarction, coronary revascularization and stroke at 1 year. Secondary endpoints will include peri-operative myocardial and acute kidney injury, intensive care unit and hospital stay, inotrope score, left ventricular ejection fraction, changes of quality of life and exercise tolerance. Patients will be randomized to receive after induction of anesthesia either RIC (4 cycles of 5 min inflation to 200 mmHg and 5 min deflation of a blood pressure cuff placed on the upper arm) or sham RIC (4 cycles of simulated inflations and deflations of the blood pressure cuff). IMPLICATIONS: The findings from the ERICCA trial have the potential to demonstrate that RIC, a simple, non-invasive and virtually cost-free intervention, can improve clinical outcomes in higher-risk patients undergoing CABG ± valve surgery.