Privacy-preserving record linkage across disparate institutions and datasets to enable a learning health system: The national COVID cohort collaborative (N3C) experience

Introduction

Research driven by real-world clinical data is increasingly vital to enabling learning health systems, but integrating such data from across disparate health systems is challenging. As part of the NCATS National COVID Cohort Collaborative (N3C), the N3C Data Enclave was established as a centralized repository of deidentified and harmonized COVID-19 patient data from institutions across the US. However, making this data most useful for research requires linking it with information such as mortality data, images, and viral variants. The objective of this project was to establish privacy-preserving record linkage (PPRL) methods to ensure that patient-level EHR data remains secure and private when governance-approved linkages with other datasets occur.

Methods

Separate agreements and approval processes govern N3C data contribution and data access. The Linkage Honest Broker (LHB), an independent neutral party (the Regenstrief Institute), ensures data linkages are robust and secure by adding an extra layer of separation between protected health information and clinical data. The LHB's PPRL methods (including algorithms, processes, and governance) match patient records using "deidentified tokens," which are hashed combinations of identifier fields that define a match across data repositories without using patients' clear-text identifiers.

Results

These methods enable three linkage functions: Deduplication, Linking Multiple Datasets, and Cohort Discovery. To date, two external repositories have been cross-linked. As of March 1, 2023, 43 sites have signed the LHB Agreement; 35 sites have sent tokens generated for 9 528 998 patients. In this initial cohort, the LHB identified 135 037 matches and 68 596 duplicates.

Conclusion

This large-scale linkage study using deidentified datasets of varying characteristics established secure methods for protecting the privacy of N3C patient data when linked for research purposes. This technology has potential for use with registries for other diseases and conditions.

Assessing the impact of privacy-preserving record linkage on record overlap and patient demographic and clinical characteristics in PCORnet®, the National Patient-Centered Clinical Research Network

OBJECTIVE

This article describes the implementation of a privacy-preserving record linkage (PPRL) solution across PCORnet®, the National Patient-Centered Clinical Research Network.

MATERIAL AND METHODS

Using a PPRL solution from Datavant, we quantified the degree of patient overlap across the network and report a de-duplicated analysis of the demographic and clinical characteristics of the PCORnet population.

RESULTS

There were ∼170M patient records across the responding Network Partners, with ∼138M (81%) of those corresponding to a unique patient. 82.1% of patients were found in a single partner and 14.7% were in 2. The percentage overlap between Partners ranged between 0% and 80% with a median of 0%. Linking patients' electronic health records with claims increased disease prevalence in every clinical characteristic, ranging between 63% and 173%.

DISCUSSION

The overlap between Partners was variable and depended on timeframe. However, patient data linkage changed the prevalence profile of the PCORnet patient population.

CONCLUSIONS

This project was one of the largest linkage efforts of its kind and demonstrates the potential value of record linkage. Linkage between Partners may be most useful in cases where there is geographic proximity between Partners, an expectation that potential linkage Partners will be able to fill gaps in data, or a longer study timeframe.

Establishing a framework for privacy-preserving record linkage among electronic health record and administrative claims databases within PCORnet®, the National Patient-Centered Clinical Research Network

OBJECTIVE

The aim of this study was to determine whether a secure, privacy-preserving record linkage (PPRL) methodology can be implemented in a scalable manner for use in a large national clinical research network.

RESULTS

We established the governance and technical capacity to support the use of PPRL across the National Patient-Centered Clinical Research Network (PCORnet^®). As a pilot, four sites used the Datavant software to transform patient personally identifiable information (PII) into de-identified tokens. We queried the sites for patients with a clinical encounter in 2018 or 2019 and matched their tokens to determine whether overlap existed. We described patient overlap among the sites and generated a "deduplicated" table of patient demographic characteristics. Overlapping patients were found in 3 of the 6 site-pairs. Following deduplication, the total patient count was 3,108,515 (0.11% reduction), with the largest reduction in count for patients with an "Other/Missing" value for Sex; from 198 to 163 (17.6% reduction). The PPRL solution successfully links patients across data sources using distributed queries without directly accessing patient PII. The overlap queries and analysis performed in this pilot is being replicated across the full network to provide additional insight into patient linkages among a distributed research network.

A Novel Patient Recruitment Strategy: Patient Selection Directly from the Community through Linkage to Clinical Data

OBJECTIVE

This article presents and describes our methods in developing a novel strategy for recruitment of underrepresented, community-based participants, for pragmatic research studies leveraging routinely collected electronic health record (EHR) data.

METHODS

We designed a new approach for recruiting eligible patients from the community, while also leveraging affiliated health systems to extract clinical data for community participants. The strategy involves methods for data collection, linkage, and tracking. In this workflow, potential participants are identified in the community and surveyed regarding eligibility. These data are then encrypted and deidentified via a hashing algorithm for linkage of the community participant back to a record at a clinical site. The linkage allows for eligibility verification and automated follow-up. Longitudinal data are collected by querying the EHR data and surveying the community participant directly. We discuss this strategy within the context of two national research projects, a clinical trial and an observational cohort study.

CONCLUSION

The community-based recruitment strategy is a novel, low-touch, clinical trial enrollment method to engage a diverse set of participants. Direct outreach to community participants, while utilizing EHR data for clinical information and follow-up, allows for efficient recruitment and follow-up strategies. This new strategy for recruitment links data reported from community participants to clinical data in the EHR and allows for eligibility verification and automated follow-up. The workflow has the potential to improve recruitment efficiency and engage traditionally underrepresented individuals in research.

From ICU to Emergency Department: 9-Year Experience with Non-Invasive Ventilation for COPD

Introduction

Non-invasive ventilation (NIV) has been shown to be beneficial for chronic obstructive pulmonary disease (COPD) patients with persistent respiratory acidosis during acute exacerbations. This clinical study described the experience of implementing an NIV program in the emergency department for COPD patients.

Methods

In the pre-implementation phase, patients who presented to the emergency department were transferred to the intensive care unit for NIV. Following the NIV program, patients had NIV commenced in the emergency department. We reported the change in hospital outcomes pre and post-implementation.

Results

A total of 153 patients received NIV, 34 in the pre-implementation phase and 119 patients in the post-implementation phase. The mean pH was 7.22±0.07 and the Acute Physiology and Chronic Health Evaluation (APACHE) II score was 18.9±4. Hospital mortality was lower in the post-implementation phase (1.7% versus 11.8%; p=0.008). The median door-to-NIV time was shorter in the post-implementation phase (64 minutes; interquartile range [IQR] 35-113) versus (457 minutes; IQR 143-1320).

Conclusions

NIV program in the emergency department is feasible and is associated with better hospital outcomes in patients with COPD. (Hong Kong j.emerg.med. 2014;21:140-147)

Patient-Centered Outcomes Research in Practice: The CAPriCORN Infrastructure

CAPriCORN, the Chicago Area Patient Centered Outcomes Research Network, is one of the eleven PCORI-funded Clinical Data Research Networks. A collaboration of six academic medical centers, a Chicago public hospital, two VA hospitals and a network of federally qualified health centers, CAPriCORN addresses the needs of a diverse community and overlapping populations. To capture complete medical records without compromising patient privacy and confidentiality, the network created policies and mechanisms for patient consultation, central IRB approval, de-identification, de-duplication, and integration of patient data by study cohort, randomization and sampling, re-identification for consent by providers and patients, and communication with patients to elicit patient-reported outcomes through validated instruments. The paper describes these policies and mechanisms and discusses two case studies to prove the feasibility and effectiveness of the network.

The impact of clinical protocols in the management of severe sepsis: a prospective cohort study

This study aimed to assess the availability of clinical protocols and their effect on compliance to the Surviving Sepsis Campaign bundles and on mortality in severe sepsis in ten Singaporean adult teaching intensive care units (ICU). The presence of 11 protocols in the ICUs, steps taken based on the Johns Hopkins University Quality and Safety Research Group's model to translate protocols into practice, and organisational characteristics were assessed. Clinical and research personnel recorded characteristics of patients with severe sepsis who were admitted in July 2009, the achievement of sepsis bundle targets and outcomes. Hospital mortality was 39% for 128 patients. Fewer than half of the ICUs had protocols for early goal-directed therapy, blood cultures, antibiotics, steroids, lung-protective ventilation and weaning. Compliance rates with the resuscitation and management bundles were 18 and 3% respectively. Units with protocols were generally not more likely to achieve associated bundle targets. Steps from the Johns Hopkins model to measure performance, engage teams and sustain and extend interventions were taken in fewer than half of the available protocols. However, on logistic regression analysis, the number of protocols available per ICU was independently and inversely associated with mortality. In conclusion, clinical protocols are infrequently available in Singapore's ICUs and when present do not generally improve compliance to the sepsis bundles. These protocols may, however, be a surrogate marker of the quality of care as they are independently associated with decreased mortality. The use of an integrated and multifaceted approach to translate protocols into practice should be considered.

Utility of the electronic information resource UpToDate for clinical decision-making at bedside rounds

INTRODUCTION

Clinical questions often arise at daily hospital bedside rounds. Yet, little information exists on how the search for answers may be facilitated. The aim of this prospective study was, therefore, to evaluate the overall utility, including the feasibility and usefulness of incorporating searches of UpToDate, a popular online information resource, into rounds.

METHODS

Doctors searched UpToDate for any unresolved clinical questions during rounds for patients in general medicine and respiratory wards, and in the medical intensive care unit of a tertiary teaching hospital. The nature of the questions and the results of the searches were recorded. Searches were deemed feasible if they were completed during the rounds and useful if they provided a satisfactory answer.

RESULTS

A total of 157 UpToDate searches were performed during the study period. Questions were raised by all ranks of clinicians from junior doctors to consultants. The searches were feasible and performed immediately during rounds 44% of the time. Each search took a median of three minutes (first quartile: two minutes, third quartile: five minutes). UpToDate provided a useful and satisfactory answer 75% of the time, a partial answer 17% of the time and no answer 9% of the time. It led to a change in investigations, diagnosis or management 37% of the time, confirmed what was originally known or planned 38% of the time and had no effect 25% of the time.

CONCLUSION

Incorporating UpToDate searches into daily bedside rounds was feasible and useful in clinical decision-making.

Effect of a blood conservation device in patients with preserved admission haemoglobin in the intensive care unit

An important iatrogenic cause of anaemia in the intensive care unit is loss of the discarded blood during phlebotomy via indwelling vascular catheters. A closed system blood conservation device has previously been shown to reduce the need for blood transfusion and to blunt the decrease of haemoglobin in intensive care unit patients. However such a device may not benefit patients who are admitted with a relatively preserved haemoglobin. In this sub-group analysis of a before-and-after study, 128 patients had admission haemoglobin > or =115 g/l and did not receive any blood transfusions while in the intensive care unit. In the control group of 50 patients a blood conservation device was not used, while in the active group of 78 patients the device was used. Use of the blood conservation device did not affect the haemoglobin trends when both groups were compared using the general linear model. For patients with admission haemoglobin > or = 115 g/l, use of a blood conservation device does not affect the subsequent rate of haemoglobin decline in the intensive care unit. These patients are unlikely to benefit from the use of such devices.

Monitoring of extubated patients: are routine arterial blood gas measurements useful and how long should patients be monitored in the intensive care unit?

Restitution of respiratory support, which may include continuous positive airway pressure, non-invasive ventilation or reintubation, is needed in some patients post-extubation. We aimed to investigate whether serial arterial blood gas measurements done in the post-extubation period would help to identify such patients and to delineate the optimal post-extubation duration for close monitoring. We retrospectively analysed 115 consecutive adult patients who were extubated following successful spontaneous breathing trials in the medical intensive care unit, excluding patients who were extubated to immediate non-invasive ventilation. Arterial blood gases were measured at one and three hours post-extubation and patients were followed for any restitution of respiratory support for the remainder of their hospital stay. Restitution of respiratory support was required for 22 of 115 (19.1%) patients, of whom 20 were originally intubated for pneumonia. These patients could all be detected clinically from deteriorating pulse oximetry or increasing drowsiness. Performing serial arterial blood gas measurements following extubation did not improve the detection rate or allow earlier detection of patient deterioration. Among the patients with pneumonia, restitution of respiratory support was required within 24 hours of extubation for 16 patients (80%) and after more than 49 hours for four patients. Serial arterial blood gas measurements at one and three hours after a planned extubation are not useful and patients originally intubated for pneumonia should be monitored post-extubation for at least 24 hours in the intensive care unit.

How well do serum sTREM-1 measurements prognosticate in septic shock?

The soluble triggering receptor expressed on myeloid cells (sTREM)-1 has emerged as a potentially useful biomarker for the diagnosis of sepsis. This study aimed to evaluate the prognostic utility of serum sTREM-1 in septic shock, in comparison with that of procalcitonin measurements. Thirty-one consecutive patients in a tertiary medical intensive care unit with septic shock were studied. sTREM-1 levels in blood were measured using a modified immunoblot array technique on days one to three of intensive care unit admission. Serum procalcitonin and interleukin (IL)-1beta, IL-6, IL-IO and tumour necrosis factor-alpha levels were also measured. No significant difference was observed in the sTREM-1 levels on the first three days between survivors and nonsurvivors. sTREM-1 levels moderately correlated with the Acute Physiology and Chronic Health Evaluation II and Sequential Organ Failure Assessment scores on day three, but did not correlate with vasopressor requirements, cytokine levels and the presence of bacteraemia. In contrast, procalcitonin levels were significantly higher in nonsurvivors than in survivors on days two and three. A significant relationship also existed between procalcitonin levels and the other variables. In conclusion, this study found that the prognostic utility of serum sTREM-1 in septic shock is poor and that procalcitonin measurements perform better in this regard.

Soluble triggering receptor expressed on myeloid cells-1 in acute respiratory infections

Levels of the soluble form of the triggering receptor expressed on myeloid cells (sTREM)-1 are elevated in severe sepsis. However, it is not known whether sTREM-1 measurements can distinguish milder bacterial infections from noninfectious inflammation. The present authors studied whether serum sTREM-1 levels differ in community-acquired pneumonia, exacerbations of chronic obstructive pulmonary disease (COPD), asthma and controls, and whether sTREM-1 may be used as a surrogate marker for the need for antibiotics. Serum sTREM-1 levels in 150 patients with pneumonia, COPD and asthma exacerbations and 62 healthy controls were measured. Serum sTREM-1 levels were significantly elevated in pneumonia (median 295.2 ng x mL(-1)), COPD (280.3 ng x mL(-1)) and asthma exacerbations (184.0 ng x mL(-1)) compared with controls (83.1 ng x mL(-1)). Levels were higher in pneumonia and Anthonisen type 1 COPD exacerbations than in type 2 and 3 COPD and asthma exacerbations. The area under the receiver operating characteristics curve for sTREM-1 as a surrogate marker for the need for antibiotics was 0.77. Serum levels of the soluble form of the triggering receptor expressed on myeloid cells-1 were elevated predominantly in pneumonia and Anthonisen type 1 COPD exacerbations versus type 2 and 3 chronic obstructive pulmonary disease exacerbations, asthma and controls. Serum levels of the soluble form of the triggering receptor expressed on myeloid cells-1 has moderate but insufficient accuracy as a surrogate marker for the need for antibiotics in lower respiratory tract infections.

Severe Sepsis and Septic Shock in Adult Patients: An Approach to Management and Future Trends

Severe sepsis is sepsis associated with acute organ dysfunction. Septic shock in turn, implies severe sepsis that has led to circulatory shock refractory to fluid resuscitation alone. The immediate approach to severe sepsis follows the ABCs of resuscitation: Airway, Breathing, and Circulation. Special emphasis on the circulation involves early goal-directed therapy, adequate fluid resuscitation, and vasopressor/inotropic support. Once the patient's cardiorespiratory status is stabilized, efforts must be directed at uncovering the source and empirically yet accurately treating the infective underpinnings of severe sepsis. Following that, each of the patient's other organ systems at risk needs to be addressed: Renal/metabolic, gastrointestinal, hematological, and endocrine. Novel treatments will target both the proinflammatory and procoagulation cascades of sepsis.