Need for clinical research informatics


Niladri Majumder
Project Analyst
Sidra Medical & Research Centre

Healthcare providers having a ‘Research Centre/Institute’ tag suffixed to their display name has been a widespread phenomenon in our country. But in reality only a handful of them actually engage in active clinical research while the others often mistakenly use this to reap commercial benefits. The fact that clinical research is the lynchpin that connects innovative technologies from basic discovery research to their application as breakthroughs in patient care is deeply undervalued by healthcare centres. Most care providers treat clinical research as a byproduct of clinical care and the former is more of an afterthought/ add on to the later.

The flow of research from bench to bedside and back again, through review in practice, necessitates a comprehensive research strategy aligned to the organisational vision. Many organisations falter with their research programmes in absence of an appropriate strategy and leadership commitment.

Developing the right informatics infrastructure through provisioning relevant IT solutions is a vital cog to a comprehensive research strategy. Moreover, with the correct business model healthcare institutions derive additional research value from a sophisticated IT infrastructure.

But improvements through the use of electronic information exchange have been slow in clinical research studies for many reasons, including the lack of informatics infrastructure, exemplified by low electronic medical records (EMR) adoption, inconsistent data standards and database architecture, and insufficient analytic tools. Even though our healthcare leaders are being increasingly aware of the need for digitising the hospital operations, potential efficiencies available through IT solutions for clinical research are largely undermined.

Studies indicated that over 75 per cent of information obtained for support of clinical trials was entered on paper[1]. Use of electronic solutions can reduce the cost of data collection by 55 per cent over paper[2]. Furthermore, the information once collected is typically entered for various needs from four-seven times by clinicians.

Additionally, three main fields in science and medicine that are currently disconnected:

1) basic research, which tries to understand the fundamental principles and phenomena that drive cells, organisms and systems in both normal and pathological conditions (such as cancer); 2) translational research and applied medicine, which represent the application of basic research to solve specific problems, aid in diseases and help society at different levels as well as
3) EMRs that have been developed as a new technology to facilitate both patient care and research by collecting and archiving patients’ history. The key bottleneck is how to efficiently integrate these three independent “parts” of medical and scientific areas in a single solution to improve patient care.

A robust clinical research informatics platform enables:

  • Integration of patient and research data
  • Sponsors to understand the progress of their research projects
  • Institutions to ensure ethical and regulatory compliance
  • Robust data collection, processing and reporting
  • Capturing and leveraging IP generated
  • Enable collaboration among partners and others in research fraternity

The clinical research workflow

Clinical research study might be stated as an overarching terminology which envisages sponsored or investigator initiated clinical trials or other survey-oriented non interventional studies. A typical clinical research study lifecycle consists of the following steps. The study originates from a hypothesis, and then is detailed in a study protocol. The protocol documentation includes study design and operational details such as the duration of the study, type of participants, inclusion and exclusion criteria for subjects, patients’ schedule for assessment and interventions, medications and dosages. Following this, the protocol is assessed by an Institutional Ethics Committee (IEC) or Institutional Review Board (IRB) to ensure the appropriateness of the clinical trial protocol as well as weigh the risks and benefits to study participants. It also reviews all study-related materials before and during the trial.

After the study is approved and site selection done, patient/subject are screened and recruited based on pre-determined inclusions and exclusion criteria. Thereafter all enrolled study subjects/patients undergo series of assessment, investigations/interventions in compliance to the protocol schedule. The clinical research staff ensures appropriate documentation, de-identification of data per relevant standards and archiving of data. Collected data are analysed and interpreted by researchers to derive inferences which form the basis for biomedical discoveries.

At the heart of clinical research is the immense data collection and analysis that determines the efficacy and safety of medical therapies. Currently, the processes for identifying subjects eligible for research, collecting study parameters, assembling information from multiple study sites, conducting oversight of study protocols, and analysing results involve manual operations which are time consuming, labour intensive, and expensive.

Capabilities of an integrated clinical research informatics model

Historically, research and healthcare information technology systems have been disconnected, supporting separate, but sometimes redundant, processes and workflows. Unfortunately, the use of disparate systems can result in patient safety concerns, inefficient processes, data quality issues and challenges.

Most Clinical Research Organizations (CROs’) and pharmaceutical companies engaged in clinical trials rely on commercial off-the-shelf Clinical Trial Management System (CTMS) solutions for digitising the trial process. A CTMS package is an integrated suite of applications sharing a common database designed to help manage clinical trails activities at different levels. But a hospital-based clinical research presents unique informatics requirements that are amenable to solutions supported by EMR systems which might be already deployed. Such scenarios solicit an EMR integrated comprehensive informatics solution which extends beyond the scope of a traditional CTMS solution. The solution should enable:

  • Proactive identification of potential research subjects from the EMR database
  • Help screen and recruit research participants
  • Research data collection through electronic data capture methods (EDC) from the EMR including web, hand held devices, and phone-based Interactive Voice Recognition (IVR)
  • IRB document management, amendment and milestone tracking
  • Randomisation and blinding of participants in a randomised control trial
  • De-identification of data according to HIPAA standards
  • Trail reporting and identify data queries that needs to be addressed
  • Adverse events reporting
  • Easily move valid data from EMR into research registries
  • Facilitate secure EMR access for research auditors/ monitors
  • Appropriately billing for research visits to the sponsor unlike the normal visit
  • Tagging hospital patients enrolled in research studies for easy identification
  • Capture rich structured data from the EMR (phenotypic) and combine with bio-informatics data (genotypic)
  • Enable export data in format (for eg: CDISC )for initial statistical purposes or downstream integration with other tools
  • Provide coding methods for fields such as pathology or medication data

The clinical research informatics model

The desired capabilities of an EMR integrated clinical research informatics solution, as discussed in the previous section, has been mapped to the enabling core IT systems functionalities. More often all these functionalities might not be met by a single software solution but an integration of multiple systems, which might be necessary to achieve the desired research outcomes. This model also might form a basis for negotiations with the EMR vendor to include additional functionalities in the solution package to support clinical research.

Moreover, this lays a long-term foundation for accelerated discovery, extensive outcomes research, and ultimately a ‘learning health system’ in which a ‘bench-to-bedside-to-bench’ cycle of information will support continual improvement in knowledge, care and health.

The model is split in two distinct zones meant for patient care and clinical research intersected by a pseudonymisation wall.

Research specific functionalities within EMR

The EMR is the centerpiece to all patient care activities in the hospital. It is capable of capturing and archiving all information (for eg: charts, orders, results, medications, diagnosis, interventions etc) generated during patient care. For a hospital engaged in clinical research, EMRs can be optimised to envisage pro research functionalities like:

  • Study feasibility and screening – The EMR has potential to be an ideal system to search for patients that are eligible for an ongoing or potential study. All patient records from the EMR is maintained in a EMR Data Warehouse (EMR-DW), a system that enables the patient records to be stored and used for analytical purposes. With advanced search capabilities, it’s also possible to retrieve a series or records to retrospective chart reviews.
  • Recruitment alerting – Inclusion criteria for clinical research studies can be programmed into the EMR, which can proactively alert the attending physician that a patient is eligible for enrolment into a study when they are in the clinic. Failure to recruit a sufficient number of eligible subjects presents a major impediment to the success of clinical trials. This can be addressed if the EMR has the capability of a real time Clinical Trial Alert (CTA). Without the assistance of IT solutions, recruitment is extremely slow, expensive, and low-yield.
  • Patient tagging and participant tracking – Identification of hospital patients involved in a research is crucial to protecting the safety and rights of participants. The EMR should be capable of flagging (a sort of identification on the patient’s record) the patient enrolled in a clinical research study with information about clinical trials/studies in which he/she is participating. This would help clinical staff, investigators, study coordinators, clinicians, and oversight bodies such as the IRB to follow participants throughout the research process and ensure that their safety and rights are protected. An integrated participant tracking enables better management and eliminates need for multiple data entry
  • Protocol document management – The system should have document management capabilities for support grants, peer review, IRB continuing review, ethics approval, adverse event reporting, etc and support the informed consent and re-consenting processes. Organisations of a larger scale prefer dedicated Enterprise Content Management (ECM) systems to fulfill this need.

Research study management system: Core functionalities

eCRF & Electronic Data Capture: Traditionally paper based Case Report Forms (CRF) have been used to transcribe relevant data manually from the medical record of a clinical research study subject. In a paper system, data are entered first on the clinical report form and second by the data entry group into an electronic system. An eCRF is a pre configured electronic form that eliminates manual data entry/re entry and fetches the data from the medical records through electronic data capture.

The Electronic Data Capture (EDC) is a system that electronically transcribes clinical data from the EMR into the electronic case report form (eCRF). EDC replaces the traditional paper-based data collection methodology to streamline data collection and management.

Since data are entered into a data collection tool only once, processing time for data entry is reduced, and transcription errors are less likely. EDC can help clean and lock data faster than traditional paper CRF systems.

Pseudonymisation: Research studies must use de-identified patient data according to relevant standards (for eg: HIPAA). This is not only a legal requirement but is essential for protecting patients’ rights. With that in effect, the patient data flow from clinical to research domain should channel through a data pseudonymisation system. This system will enable de-identification of patient identifiable data to accepted standards such as HIPPA, assign codes to the data as well as support re-identification of patients in case of medical emergencies. Following pseudonymisation, the de-identified data is expected to reside in a separate research domain where can be used by researchers who cannot identify which patient the data is particular to. This includes having the capability to handle cases of very small populations.

The de-identification system needs to integrate with any system where Personal Health Information (PHI) is collected and will then be used for research purposes.

Randomisation and blinding: Randomisation capability enables locally sponsored interventional clinical studies for systematic assigning patients to different study arms in a statistically robust fashion. It needs to interface with clinical systems including, but not limited to EMR, the pharmacy system, and the research systems and its associated electronic data management systems.

For blinded clinical trial that involves providing a blinded drug-based intervention the hospital pharmacy system needs capabilities to un-blind in an emergency the study treatments and to manage the resupply to the patient on the behalf of the study.

Data management: Accurate, reliable, usable data — it’s the lifeblood of clinical research. The method an investigator chooses for collecting, storing and analysing data can mean the difference between a study that advances the science and improves patient care, or a study with inconclusive results and no publications. Typically the data management functionality will capture all of the structured and unstructured data that is captured in study documents. This can be used to provide a warehouse of all research data that has been captured from the research activities. Also there should be a role and site level access, which allows users from other institutions to contribute data to the studies.

The system should not only be able to support studies requiring low to medium data collection complexity but also should be able to manage regulatory compliance (for eg: 21 CFR Part 11), sophisticated data management and quality control, for multi-site trials.

Data marts and knowledge management: A project data mart is typically project specific collection of data that enables clinical and research data to be extracted from their primary data management systems and combines and curates the study data into a system that enables researchers to explore and analyse their data according to the study design.

A knowledge management system is a searchable repository of research results (or facts) that capture the published and published findings of research activities. This includes statistical relationships between patients and their biology and integrates entities across patients, literature, markers and drugs. Data collected from across the research activates needs to be analysed to make inferences about the specific research hypothesis.

The way ahead

India is being touted as one of the fastest growing market globally for clinical trials with more than 15 per cent of global clinical trials expected to be carried out in the country [3]. With increasing market size and complexity there is a need of better ways to transfer and access patient information electronically. A clinical research enterprise that is configured around such electronic information systems will yield more rapid scientific discovery and will provide significant support of related activities including: comparative effectiveness of clinical trials and other outcomes-related research; quality of care measurement; public health and safety monitoring, and post-marketing surveillance.

With a widespread EMR adoption, the interface is gradually becoming more porous and productive, while providing the required oversight. These improvements will positively affect patients, physicians, researchers, entities which invest in clinical research, and ultimately all who look to scientific discovery to propel medical advancement.

References:
[1] Alschuler L, Bain L, Kush RD. Improving data collection for patient care and clinical trials. Sci Career Mag Mar 26 2004
[2] Pavlovic I, Kern T, Miklavcic D. Comparison of paper-based and electronic data collection process in clinical trials: Costs simulation study. Contemp Clin Trials 2009; 30: 300-16.
[3] http://www.outsourcing-pharma.com/Clinical-Development/India-tipped-for-15-share-of-clinical-sector-by-2011

Literature Review:
http://www.indianexpress.com/news/crackdown-two-hospitals-lose-research-institute-tag/842869/1
http://ctri.nic.in/Clinicaltrials/login.php
www.epic.com
www.cerner.com

Comments (0)
Add Comment