Informatics and the clinical laboratory.

Informatics and the Clinical Laboratory

Review Article

Informatics and the Clinical Laboratory Richard G Jones,1 Owen A Johnson,2 Gifford Batstone3

Deputy Director, Yorkshire Centre Health Informatics, University of Leeds, Leeds, UK; 2 School of Computing, University of Leeds, Leeds, UK; 3 Brighton and Sussex University Hospitals NHS Trust, Brighton, UK. *For correspondence: Mr Owen Johnson, [email protected] 1

Abstract The nature of pathology services is changing under the combined pressures of increasing workloads, cost constraints and technological advancement. In the face of this, laboratory systems need to meet new demands for data exchange with clinical electronic record systems for test requesting and results reporting. As these needs develop, new challenges are emerging especially with respect to the format and content of the datasets which are being exchanged. If the potential for the inclusion of intelligent systems in both these areas is to be realised, the continued dialogue between clinicians and laboratory information specialists is of paramount importance. Requirements of information technology (IT) in pathology, now extend well beyond the provision of purely analytical data. With the aim of achieving seamless integration of laboratory data into the total clinical pathway, ‘Informatics’ – the art and science of turning data into useful information – is becoming increasingly important in laboratory medicine. Informatics is a powerful tool in pathology – whether in implementing processes for pathology modernisation, introducing new diagnostic modalities (e.g. proteomics, genomics), providing timely and evidence-based disease management, or enabling best use of limited and often costly resources. Providing appropriate information to empowered and interested patients – which requires critical assessment of the ever-increasing volume of information available – can also benefit greatly from appropriate use of informatics in enhancing self-management of long term conditions. The increasing demands placed on pathology information systems in the context of wider developmental change in healthcare delivery are explored in this review. General trends in medical informatics are reflected in current priorities for laboratory medicine, including the need for unified electronic records, computerised order entry, data security and recovery, and audit. We conclude that there is a need to rethink the architecture of pathology systems and in particular to address the changed environment in which electronic patient record systems are maturing rapidly. The opportunity for laboratory-based informaticians to work collaboratively with clinical systems developers to embed clinically intelligent decision support systems should not be missed.

Introduction Laboratory computing has been at the leading edge of health information systems application development for the past five decades1-3 and fully integrated laboratory information technology (IT) systems have become the de facto prerequisites for efficient clinical service delivery and laboratory management. This assertion is borne out by the observation that laboratory systems are usually the most sophisticated systems in any health organisation anywhere in the world. However the health informatics landscape is changing with the development of ever more sophisticated electronic patient record systems and pathology computing is in danger of falling behind. The fundamental models of laboratory system design were elaborated in the 1970s in projects such as the

UK NHS4 Phoenix programme, and these models are still at the core of the latest systems being installed at present (Table 1). However, new approaches to managing information are now required, in order to respond to changing demands of healthcare delivery: 1. The consolidation of pathology services into larger units of organisation, e.g. managed clinical networks of laboratories 2. The increased role for primary and community care within the broader health service 3. Introduction of new approaches to service delivery, e.g. point-of-care testing (POCT)

Clin Biochem Rev 35 (3) 2014 177

Jones RG, Johnson OA & Batstone G

Table 1. A summary of the historical developments of laboratory systems since 1960. Year

Major developments

1960 - 1970

Data reduction - analogue-digital conversion Radioimmunoassay analysis

1970 - 1980

On-line instrumentation control

1980 - 1990

Intelligent workflow control Real-time links to patient administration systems Automated quality control calculations Automated report comments

1990 - 2000

Email of reports On-line data access from wards and clinics Remote test ordering

2000 - 2010

Intelligent diagnostic decision support Integrated robotic laboratories Control of distributed bedside analysers

2010 -

EPR integration Digital histopathology and genomics Patient access to data Consolidation and large scale interoperability

4. More open access to the knowledge base supporting evidence-based medicine for clinicians, patients and public 5. Integrated management of all aspects of clinical services 6. Demands for direct patient access to laboratory results for self-management. Many new developments are also presenting laboratory IT specialists with significant challenges. These include the emergence of maturing electronic record systems, digital imaging systems, high throughput gene sequencing, companion diagnostics, budget restrictions, hospital consolidations, laboratory consolidation, growth of the Internet and consumer data access. These are all precipitating a fundamental re-evaluation of the way in which pathology services are provided. The outcome has been, and is, to cause development interest to move from internal laboratory issues, such as instrument interfacing and the invention of sophisticated data reduction algorithms, towards external services issues such as test ordering and results delivery to the newly emerging clinical systems. Historical Background The reason for early laboratory involvement in computing was the pressure of handling large volumes of data efficiently.3 Initially, systems were designed to automate data reduction 178 Clin Biochem Rev 35 (3) 2014

carrying out functions such as analogue-to-digital conversion for raw data generated from spectrophotometers and other instrumentation. Later, they were developed to handle sophisticated calculations such as the curve fitting techniques used in radioimmunoassay. However, they were also soon adapted to automated laboratory administration since the housekeeping tasks of results storage and retrieval were becoming onerous as the growth in demand for laboratory tests built up during the 1970s.3 Most general readers will not appreciate the data processing requirements of laboratories. Analysis of data from the UK suggest that an average multi-disciplinary pathology service typically handles between 1 and 3 million patient specimens per year each, generating 10–30 million reportable test results on populations ranging from 0.5 to 1.5 million. In the UK, the median laboratory repertoire of tests is 189; such a laboratory will serve approximately 500–1500 consultant clinicians and a similar number of general practitioners (GPs) within an average of 300 practices. Such operations are of industrial scale.5 It is a tribute to the systems designed and developed in the early 80s that they can handle this volume of work. Paradoxically, these early systems developed in the Massachusetts General Hospital Utility Multi-Programming System (MUMPS), a language often viewed as outdated or obsolete, continued to outperform many newcomers developed on more modern


relational-database technologies.6 With developments in technology in the last decade, this situation is now changing and the drawbacks of the 1980s languages are becoming increasing evident, particularly in regards to support for business functions and in terms of long-term maintenance. Current laboratory information management systems (LIMS) typically enable: l Recording of all requests for all tests l On-line, real-time linking of the LIMS to automated analytical instruments l Sample tracking and workflow management l Worksheet generation for manual tests l Automated validation of test results l Real-time recording of quality control data l Electronic delivery of results to clinical users l Implementation of decision support systems to enhance clinical outputs l Support of data analysis for audit, clinical risk management, disease surveillance and epidemiology [e.g. cancer registration, screening programmes, communicable disease reporting, and external quality assessment (EQA) data management] although often requiring specific informaticians input and programming. One impact of current laboratory systems has been to make the internal operations of the laboratory a relatively paper-free exercise, in marked contrast to clinical areas within hospitals and more in line with the well-developed IT systems in primary care in the UK.7,8 The current systems log all requests for tests, transmit these on-line, and in real-time, to automated analytical instruments and receive back the test results. The operation is usually controlled by use of unique bar-coded sample IDs to minimise the risks of data transcription errors. Once work is completed reports are printed ready for dispatch to wards and clinics, and in many parts of the world this reporting is now also predominantly electronic. In addition to this basic control of workflow the systems provide a multitude of other sophisticated functions. They track quality control samples and this data is a major component in the quality assurance procedures of the laboratory alongside the audit and reagent tracking capability of many analytical instrument control systems. Management data is generated as a ‘free’ by-product of the automated operational procedures; datasets captured at source are transferred electronically to financial systems and workload analysis is itself an automated and routine activity (Figure 1). In addition, the data forms a rich source of material for research and audit, especially when it is matched against datasets containing diagnostic or outcome data on individual patients. Indeed, access to this data for research is becoming increasingly recognised as a significant value-added product with the UK for example investing

Figure 1. Central role of electronic patient record to health care delivery. many millions in programmes to exploit this rich resource.9-11 Current laboratory systems also contribute to the clinical role of the departments through the use of internal rule-based procedures for highlighting exceptions and automatically generating interpretative comments on individual reports. For outsiders it is sometimes hard to appreciate the virtual manner in which this process takes place. The clinical functions take place remotely from the site of analysis, in some cases in different hospitals miles apart, and involve co-operative, multidisciplinary working between different groups of staff. Approximately 80% of results can pass through the systems without direct human intervention.12 Optimal System Organisation The informatics requirements for an integrated pathology service can usefully be considered from a business perspective in terms of four layers of operation, each having a different role in service delivery (Table 2). These four layers (Infrastructure, Operational, Business Support, Executive) must function in an integrated manner so that all services are fully supported. Seemingly trivial incompatibilities (e.g. in test code definitions) may have major adverse impacts on data transfer, whether for clinical purposes or for analysis of data for managerial purposes. This model is analogous to the developing role of the electronic patient record which supports clinical activity as a core pre-requisite while also supporting audit and knowledge generation and thus contributing to service delivery improvement (Figure 1).13 What is the Business Model of Pathology and Laboratory Medicine? In order to understand the informatics and IT requirements of a modern pathology service it is necessary to understand the underlying business models. These have been changing in recent years with less emphasis on ‘analytical’ aspects and more on ‘clinical service delivery’. Accurate, scientifically valid analysis is still important but for it to be effective it must Clin Biochem Rev 35 (3) 2014 179


Table 2. Informatics requirements for an integrated pathology service. Layer

Description

Clinical Functions

Management Functions

Information Content

4

Executive Systems

Clinical Decision Support

Performance Monitoring Planning Quality Management Research exploitation

Knowledge

3

Business Support Systems

Ward Requesting Clinical Reporting R&D Support

Workload Statistics Contracting Supplies management Quality Assurance statistics National returns

Aggregate information

2

Operational Systems (e.g. Traditional LIMS and Workstations)

Analytical Control Reporting Point-of-care testing networks

E-mail Word Processing Web Browsing

Operational Data

1

Network / Infrastructure

Servers / PCs Network / Printers/voice & image management

Operational

be fully integrated into the appropriate clinical decision loop. This was articulated well by Lundberg as the ‘Brain to Brain’ Cycle – the clinicians ask for tests to investigate a condition but the decision loop is not regarded as closed until the report has been seen, evaluated and acted upon by the clinician or clinical team.14 Indeed in many parts of the world, pathology services are being renamed as laboratory medicine to reflect the broader clinical role in the delivery of healthcare.

results in an information attribute whilst the latter results in an action. The traditional view is of laboratories as production units where the products are test results. However, in line with changes in other industries, the focus is shifting to business models based on a more customer-focussed and service-based approach where the value-chain is being identified and used to determine what should and what should not be provided (see Figure 2).15

Laboratory medicine is perceived and practiced in different ways in different countries, the perceptions reflecting the difference in importance attributed to the production of a test result and the appropriate use of a test result. The former

A laboratory medicine service thus comprises (i) the people who employ (ii) technology in analytical processes to (iii) produce diagnostic tests results, which are (iv) then, interpreted to enable clinicians to make appropriate decisions. The laboratory medicine service is not the technology, the

Figure 2. Contrasting features of traditional and service organisations. Adapted from Price & Jones.13

Figure 3. The core functions of the laboratory structured as service oriented business.

180 Clin Biochem Rev 35 (3) 2014


people or the processes per se, but the information product that leads to better outcomes for patients, for clinicians and for health care providers and commissioners. By way of analogy with other service industries such as banking, the analytical functions can be viewed as ‘back office’ activity and the clinical interface as ‘customer service’ or ‘front office’ functions (Figure 3). The use of information technology is key to this shift of emphasis and to integration into the emerging world of e-health.16 Thus the core of the laboratory medicine service lies in ensuring that the appropriate test is requested (or ‘ordered’), that the right result is produced, and that the result is understood and used correctly for the benefit of the patient i.e. to generate an improved outcome (Figure 4). The two key processes are the analytical process and the knowledge management process – the former emphasises production of the right result on the right patient, at the right time. The latter involves providing advice, where required, on the appropriate test, the frequency of testing and the meaning of the result, including the action that might be taken. After many years of investigation and many false starts evidence is now accumulating that the deployment of decision support systems can significantly improve the appropriateness of test selection.17-20 Supporting the Patient Journey Most health systems are now being optimised around the concept of a patient pathway or journey (Figure 5). In this context the analytic approach emphasises the role of the tests as they are employed to screen for disease, make or refute a diagnosis, to assess prognosis and/or to inform and follow-up treatment. By contrast, the knowledge management approach can include the choice of treatment (e.g. medical or surgical procedure, what drug to use), optimisation of the dose of drug, monitoring the efficacy of the drug, and checking patient

Figure 4. An outcomes driven framework for the laboratory medicine service. Reproduced with permission from Price & Jones.13

compliance. In both cases, information technology is the key to ensuring that laboratory information can be accessed appropriately at any point along the pathway. In addition the use of data on likelihood ratios and predictive values of tests in clinical contexts will become hugely more valuable than the current habit of flagging high and low values as a ‘low values relative to a reference range’. Key Issues for Modern Pathology Information Systems Identification of Pathology Service Needs Ideally, the information system functionality for pathology IT systems should be determined by service needs rather than drive these. In the UK, the NHS Pathology Modernisation Agenda reinforced by the Carter Review provided a major challenge for pathology informatics, requiring a significant shift in focus, away from traditional internal laboratory issues (instrument interfacing, data reduction technologies, robotics etc.), to broader external service issues such as clinical system integration, test requesting and reporting of results, to clinical audit and management.21 Knowledge management and performance management (reflected in value added clinical output) are now key drivers, superseding earlier preoccupations with workload and cost.19,22 Specifically, the trend towards managed clinical networks requires the development of new information systems integrated across different hospital trusts, multiple clinical laboratories and component disciplines. Expansion of primary care services also requires greater linking of requests and reports from multiple sites together with expanded user access, while ensuring appropriate security and confidentiality.8,23,24 Larger test repertoires and greater use of new technologies including POCT will almost certainly lead to increased service demand. Containing this demand within finite resources requires careful management, to which effective

Figure 5. Illustrating the points at which the laboratory medicine services contribute to the patient journey. Reproduced with permission from Price & Jones.13 (GP: General Practice; IP: Inpatient; OP: Outpatient.) Clin Biochem Rev 35 (3) 2014 181


IT can make a major contribution. For example, providing fully integrated electronic patient records and links to clinical decision mechanisms can support evidence-based clinical care.25,26 The transition to the concept of ‘risk-based quality control’ brings new opportunities for laboratory informatics to provide information on laboratory performance related to biological variation rather than current approaches – clinically orientated QC rather than laboratory practice based QC.27 These approaches use methods such as Six Sigma to assess laboratory performance and to characterise between machine / module variation within the laboratory.28 Availability of IT Systems to Meet New Pathology Service Needs Systems required to fully meet the objectives outlined above are not yet available, as the systems in use were designed long before these requirements emerged. Laboratories have therefore been forced to adopt piecemeal approaches typically patching together large LIMS and analytical systems with a plethora of spreadsheets, Microsoft Access databases, custom reporting tools and bespoke programming. As a result many laboratories have ended up with tens if not hundreds of systems more closely resembling a shanty town than a modern, strategic system with a coherent architecture. Planning new IT systems requires a medium-to long-term approach and in the interim, pragmatic procurement decisions may be necessary to facilitate service reconfiguration. However, it is essential that these decisions do not conflict with strategic requirements.21 The main challenge is to consider a new approach to the architecture of laboratory systems. The scale necessitates thinking about enterprise architectures in order to ensure that the multiplicity of systems can be coherently integrated. (Figure 6) illustrates, how as organisations scale, there is a

Figure 6. The scaling of system architectures to support different size organisations. 182 Clin Biochem Rev 35 (3) 2014

need for different types of systems. Small organisations can work with PC-based personal software but as they grow, there is a need to use larger more complex packages and eventually to invest in enterprise solutions bought from large corporate vendors. Laboratory IT is currently in the middle of this graph with a hybrid of large monolithic LIMS glued together with a patchwork of locally derived solutions. In many cases taking an enterprise approach will enable considerable rationalisation and consolidation of systems. The net benefits should be reduced cost through reduction in systems costs, reduced complexity and hence savings in the maintenance costs for both IT revenue and staff. A possible architecture is shown in (Figure 7) where a large corporate data warehouse is linked to a range of sub-systems through an integration engine responsible for intersystem communications. A more advanced approach might be to use a service-bus architecture and some recent systems are adopting this approach. Indeed there are examples of very large scale cloud-based solutions handling up to 17 million requests per annum for a network of more than 100 laboratories and 1000 sample collection sites.29 The architecture rethink also needs to take account of the fact that many functions traditionally taking place on the LIMS are now supported on other systems within the portfolio of systems used by many hospitals. For example, most internal Quality Assurance will be devolved to the systems supporting the analytical platforms and, with electronic order communications and reporting systems, the data entry and clinical integration of results will take place on clinical systems used outside the laboratory. There will be profound effects on decision support – an area where laboratories have struggled to provide consistent support as the information on clinical context, for example, drug therapy in thyroid disease, has been hard to capture consistently. Many auto-generated comments on laboratory results are often, from an expert clinician view, anodyne and inappropriate. Advanced clinical systems, such as those used in multidisciplinary team meetings and those

Figure 7. A schematic architecture based on the use of an integration engine and a data warehouse.


offering e-consultation support, allow reporting directly into the clinical record with the laboratory clinician or scientist being able to interrogate the relevant areas of the clinical record and make much more intelligent and relevant judgments. One example is recent interest in identifying acute and chronic kidney injury which has demonstrated the limitations of laboratory-based algorithms. Though able to detect high or changing levels of creatinine, the methods have insufficient specificity and the overall effect has been to generate an excess of false positive alerts. Moving the algorithms into the clinical systems where they can exclude already identified cases or can include drug therapy and other clinical indicators, is the best option for increasing the specificity and more effective reporting, for example: in alerts for acute kidney injury.30 One lesson from this work is that cross-discipline development of rules whether they be instantiated in clinical or laboratory systems, is needed to make best use of specialist expertise. It points to a potential new role for laboratory informaticians in working with clinical colleagues in defining optimal rule sets. Laboratory to Laboratory Communications Though the majority of core laboratory services are highly automated and computerised, the referral of samples to other centres is still mainly paper driven. It is estimated that around six million pieces of paper are sent between 152 NHS laboratories in the UK and manually input into computer systems every year, using an estimated 300 person years and introducing the risk of errors and delay in patients receiving their results.31 The volume of lab-to-lab communications worldwide will continue to grow with the consolidation of specialist services, the rapid growth in POCT and the opening up of the marketplace for laboratory services. In the UK, the National Pathology Exchange (NPEx), provides a national digital hub for the electronic transfer of these lab-to-lab communications creating a streamlined, paperless process. The NPEx service is based on direct interfaces between laboratory LIMS systems to a secure NHS data centre containing a digital messaging engine which converts proprietary interfaces to HL7 messaging and national coding standards.32 Likened to the services provided by a telephone exchange, NPEx connects all laboratories to each other so that lab-tolab communications can be easily ‘switched’ to the cheapest, fastest or most reliable supplier creating a national marketplace in pathology testing. NPEx uses a ‘software-as-a-service’ model – participating labs pay a quarterly subscription fee, which covers network connection and support. In the UK, 30 NHS laboratories have moved to NPEx with a target of 50% of the English NHS adoption by end of 2014.33 Other counties and regions have not yet developed similar national exchanges but many recognise the same challenges and are searching for similar solutions.

Factors Likely to Influence the Future Development of Laboratory IT Systems The factors that will influence the development of systems in the future will include: The Global Nature of the Diagnostics Business The laboratory IT systems market is now global. Consolidation of the diagnostics industry and the introduction of large scale robotic analytical platforms has seen companies offering complete, integrated IT platforms and indeed many specialist IT companies have been acquired by the bigger suppliers. Some niche players provide vertically integrated platforms, for example Perkin Elmer with the Lifecycle product which integrates pre- and postnatal screening functions with specialist robotic equipment to handle blood spot samples.34 Siemens, Roche and Abbott similarly can offer fully integrated packages linking robotic and information handling functions at scale though even in these cases there is room for continuing rationalisation. Procurement Policies Future IT systems may be selected as part of a wider health informatics strategy and this may reduce the scope for purchasing customised pathology systems from specialist suppliers. In several recent procurements in the UK, a strategic decision to adopt a single hospital-wide patient system inevitably also determined the new laboratory information system.21 This may reflect the fact that many current LIMS lack open and sophisticated APIs and hence offer only limited interoperability with clinical systems. With better definition of laboratory interfaces there should be no constraint on what IT systems laboratories choose to optimise their internal operations. There will also be a need to rethink the scale of procurement if the economies of scale of IT are to be realised. In England the number of laboratory organisations has fallen by 25% in the last decade with a trend towards reduction from 150 hospital-focused services towards 20-30 networked or consolidated service organisations over the next five years.5 Specialist Systems Developments For some emergent specialties new functions are required which are demanding new approaches. In clinical genetics for example, the use of Next Generation Sequencing (NGS) places huge demands on storage capacity and the need to link to analytical tools developed in bio-informatics. Genetics also has a different model of patient data and systems must be able to retain and integrate family pedigree data. Because of the sensitivity of such data and risks of inadvertent disclosure of third party information these systems demand extended information governance procedures and their integration into the enterprise-wide architecture is non-trivial.35 Similarly systems supporting transplants and blood banking use many



routine test techniques to support donor-recipient models and, in the case of blood banking, also need to track and trace blood products from donor source to the recipient. The development of bedside technologies such as barcode scanning of a patient’s wrist band to give positive patient ID and ensure product to patient matching has been a major success, helping avoid serious transfusion errors and reducing morbidity and mortality rates.36 Issues of Direct Patient Access There is increasing interest both in self-testing and in patient access to health records. In some specialist areas of practice where shared records are used, patients already have access to pathology results, e.g. obstetrics. In the UK there are currently plans to increase patient access to pathology data and in the US, health providers such as Kaiser Permanente, are supporting patient portals where millions of their clients can access laboratory data directly.37 Potential growth in this area could have significant impact on future requirements for how access electronic data is managed, governed and how laboratory results are best presented for patient use. Enabling patient access to their results demonstrates, based on the Kaiser experience, that investment in laboratory IT can generate savings in other areas of health care services and the need for whole systems thinking about informatics. Implementation of Existing and Future Standards Standards have been a core element of laboratory computing for decades particularly as regards instrument interfacing.24 Given the growth in point-of-care systems, electronic reporting and delivery of data to patient portals, secure and efficient exchange of data is fundamental to laboratory service. This can best be achieved by adopting appropriate standards at all levels, including technical (e.g. networking, hardware, software), communications, managerial (e.g. benchmarking and performance monitoring), and emerging clinical standards such as the UK National Service Frameworks (NSF) minimum data sets for Cancer or Coronary Heart Disease.38-40 Key to successful implementation is a common and consistent set of patient identifiers, most conveniently the individual’s Social Security Number or health ID number (NHS number in the UK) backed up by corroborating demographics (Name / Date of Birth) since reliance on numbers alone can be prone to errors.41 Working to ensure that a common patient identifier is used is a key objective of many health systems and the GS1 organisation responsible for bar-code standards in Europe has developed a strong standard in this area including specifications for such items as patient wrist bands.42


Figure 8. Graphical representation of the connectedness of PMIP across English laboratories and GP primary care units. The prominent nodes are the laboratories and the fringes are GP recipients. Note that in many case multiple laboratories are connected to a single GP unit. Coding Standards Coding was recognised in the 1969 paper of Whitehead et al as of key concern to laboratories.3 SNOMED, the current leading global clinical coding and terminology standard, itself emerged from pathology being developed and fostered by the College of American Pathologists (CAP).43 There is a need however to consider the current state of standards in the laboratory domain as it is becoming evident that those in use are not fit-for-purpose. The reason for this is the increasing need to support semantic interoperability between systems to ensure that clinical communications are safe and effective. Robust and reliable information and IT are required to support communication between geographically distant laboratories, including rapid communication of both test requests and test results, e.g. between GPs, hospital clinicians and pathology laboratories, convenient access to advice from specialist staff and specialist testing facilities located elsewhere as well as appropriate access to patient information, test protocols, and supporting evidence-based information and references. To be efficient this information exchange needs to be supported and indexed coherently through an unambiguous taxonomy. In the UK there has been increasing concern that the currently used coding standard (Read codes) is no longer fit-forpurpose. Read codes are based on a clinical terminology and are pre-coordinated with one code indicating a single clinical concept, e.g. serum sodium concentration, or plasma sodium concentration.44 In the UK a nationally agreed subset of Read codes, the Pathology Bounded Code List, is used for electronic reporting between laboratories and general


practice. However, since laboratories are free to report results using units of measurement of their own choosing for many analytes, e.g. PTH, troponin, Ca125, it is still possible for clinical systems to receive data which (unless transformed) is incompatible. These codes form the basis for the Pathology Messaging Implementation Project (PMIP), the national NHS system which connects up to 120 of 150 UK laboratories feeds into a single clinical records system. The degree of interconnectedness which is supported by the UK national reporting infrastructure is illustrated in (Figure 8) and the trend is for this to increase. This interconnectedness between laboratories and care providers into complex national networks introduces new risks. There are now many reports of clinical errors as clinicians may not appreciate the need for the transformations or may become confused when they see different units on single reports, and are significantly more difficult to interpret than those used for tabular or graphical displays where data is normalised to a standard unit of measurement.5 Work is therefore in progress in the UK to achieve consensus on appropriate units of measurement and to begin to enforce the use of such units nationally.33 The other major global pathology coding standard is LOINC.45 LOINC is also pre-coordinated and does have a more nuanced and structured approach than Read since there is an extended definition of the codes and associated attributes including units of measurement. However it still suffers the same drawbacks as Read, namely combinatorial explosion since each specimen type for a particular analyte requires a new code and for time-based samples, each time point and each specimen type needs another code. At the last count the Read reporting set for the UK pathology comprised 3,700 codes and LOINC has more than 70,000 codes. At this scale code-sets become unmanageable and examination of the 136 LOINC codes for c-peptide reveals serious inconsistencies in the naming conventions and similar unit of measurement conflicts to those described above. A particular example about to be introduced is LOINC 74722-0 “At the time of the event, the device was placed within the patient’s tissue” which demonstrates the impurity of the system which will confound future attempts to use this for systematic analysis. Furthermore, both LOINC and the Read codes use the same codes for resulting and requesting and have a limited scope, being most highly developed for clinical chemistry, serology and haematology. UK National Laboratory Medicine Catalogue (NLMC) The response in the UK has been a programme to develop a new National Laboratory Medicine Catalogue (NLMC) based on relational, post-coordinated principles covering not only the reporting standards but also those needed for test requesting and extending beyond the numeric disciplines into microbiology, histopathology and genetics (Figure 9).46

Figure 9. The component objects of National Laboratory Medicine Catalogue. The NLMC provides a logical sequence from request to report with all the essential information for interpretation of the report. Obviously not all the items are required for all requests or reports. The data combination indicator is a scoring system to indicate the safety of combining results on a laboratory analyte when those results are from different laboratories which might be using incompatible methodologies. As NLMC is SNOMED coded, the data may be used in other electronic communications such as discharge summaries and clinic letters thereby supporting interoperability between laboratory and clinical systems. Other Initiatives The need to standardise terminologies and use of units of measurement is being widely recognised. In Australia the Royal College of Pathologists of Australia is supporting a major project similar to the NLMC which is attempting to standardise the use of SNOMED-CT, LOINC, UCUM and HL7 across Australian laboratories.47 Knowledge Management The principal driver for IT in pathology is to ensure that pathology services are used to best effect for patient care and to improve the health of the population at large. Pathology IT systems can support and promote best practice, both within laboratories and by external users but as yet the evidence base is immature.13,48-50 Supporting Best Practice Clinical guidelines or protocols have been in limited use for many years, but the current emphasis on evidence-based medical practice strongly encourages increased involvement of all staff in their design, maintenance and use. Local dissemination and maintenance of protocols and guidelines is essential, together with training in guideline development and deployment. IT provides a powerful means of promoting



and integrating high standards of laboratory practice within the wider clinical environment. Widespread implementation of electronic patient records should significantly improve the way in which pathology services are used, through better demand management, targeted requesting and more effective use of results. Electronic pathology order requesting is now well established and is becoming the norm in many centres. This technology potentially should facilitate the normal checks performed in the laboratory to monitor requests and if used well, can improve the appropriateness of test requesting. The current generation of systems however are largely based on third party middleware software which has limited access to clinical data and whose rule-sets are therefore limited to controlling repeat testing to prevent repeats at too frequent intervals, ensuring tests are requested at appropriate times after drug doses or in reflex requesting combinations of tests such as electrolytes when digoxin is monitored. As electronic record systems evolve it is likely the requesting will become a native function of these systems such that proactive requesting can occur driven by clinical events and information including the scheduling of tests at optimum potentially individualised intervals. As an example, evidence is emerging from a detailed outcome analysis of the use of HbA1c in diabetes that the optimum interval between tests is 16 weeks. Incorporating such rules into electronic record systems would be relatively easy and the system could be set up so as to automate the patient contact and follow-up clinic schedules.51 In considering the integration with electronic record systems, the following requirements should therefore be taken into account to encourage appropriate test requesting: l The requester should have access to the appropriate evidence base at the time tests are requested, including agreed local protocols and guidelines, and should also have access to external resources on best practice. Phlebotomists and junior doctors require information about specimen type, sample timings etc. in order to ensure the quality of the pathology service offered and optimise its use. Hospital Intranets are increasingly being used for this purpose, and there have been some significant successes particularly in microbiology and infection control.52 l Ordering/requesting systems need to be designed in a fashion that ensures the necessary information is available to the laboratory to enable the most appropriate test to be undertaken. This is particularly important in disciplines such as microbiology where the apparent variation in the incidence of infections may be due as much to laboratory decisions on which plates to use as true differences in


infection rates. The use of Standardised Microbiological Investigation protocols in the UK will only function effectively if the links between clinical context and laboratory procedures are effective.53 l Clinical systems will need to be developed so as to ‘share’ information. For example, a pharmacy system might have rules which require access to the pathology system to check serum creatinine and potassium concentrations if digoxin is prescribed. Agreement between pathology and other clinical specialties regarding clinical alerts will also be required. l The decision to request a test needs to be made in the full knowledge of tests previously performed. Data should be displayed so as to highlight trends, eliminate unnecessary repeat requests and draw attention to alert values. l The system should be cross-referenced to clinical care pathways and should prompt requesting of additional appropriate tests if not ordered. Exploiting Decision Support Systems in Pathology Increasingly sophisticated, rules-based systems are being used in pathology particularly in the areas of test selection, result validation, image analysis, clinical reporting and risk calculation (e.g. Downs Screening). It is essential that such systems are appropriately quality assured, with particular attention to liability and risk-management.54,55 The potential risks of inappropriate decision-making using such systems require the development of robust processes for their selection, implementation and management. These must take account of the developing legal regulations for software used in medical devices such as the European Union In vitro Diagnostics Directive.56 Ensuring that clinical users understand the implications of inappropriate and unregulated use of these systems is also essential. Whilst decision support tools are becoming regulated there will be increasing reliance on bioinformatics tools in laboratories and attention should be given to how these are validated, regulated and version controlled. Supporting and Exploiting Research and Development in Informatics The core elements of this service are supported by a range of complementary activities that are founded on the research that generates evidence on the utility and effectiveness of a test and the dissemination of that evidence. If technology is to be used for testing at the point-of-care, then education, training and quality assurance are aspects of the overall service, as it is assumed that the use of POCT will complement an underlying dependence on the laboratory service.


Epidemiological analysis of the results of tests performed in clinical laboratories is essential to improving patient care and enhancing our understanding of disease-related processes. Pathology IT systems should provide support for these activities. Ensuring that the laboratory service is delivered and used properly is a question that has taxed many health professionals, including providers, users, purchasers and commissioners, for many years. This includes the practical issues associated with logistics and communications – from order communications, through specimen collection and transportation to result communication.57 Evidence shows that if these are managed separately then there is a far greater risk of fragmentation of services. The increasing integration of clinical data in the future will raise many new problems of data classification and standardisation, and greater understanding of the syntax and semantics of information content will be required to solve them. These skills will relate more to linguistics and logic than to electronics, and early investment in training and development of suitable individuals will be highly advantageous. Throughout the world programmes of multidisciplinary research are already addressing the following topics: l Decision support l System usability l Long term storage & retrieval l Data ‘mining’ – clinical knowledge engineering l Data and information standards Research and Development In some cases laboratory data may contribute to development of new database resources of regional, national or international interest (e.g. the UK Association of Clinical Biochemists’ AssayFinder58). Specialist services have access to large databases of information linked to the specific analytical services they offer. They are in a unique position to contribute national and international data repositories supporting the practice of evidence-based pathology. The availability of such information is made available for online publication through a pathology branch library of the National Institute for Health and Care Excellence.59 Supporting Communications Current trends in technology are blurring the boundaries between computers and communication systems. Networked, handheld devices and services such as text messaging provide major opportunities for rethinking how information is communicated within the pathology service and between the pathology service and its users. This will become a major role for laboratory services as patient self-management of

chronic conditions becomes the norm as, for example in diabetes and anticoagulant monitoring.60 These are examples of disintermediation.61 Disintermediation The concept of disintermediation in pathology is the ability to use pathology data to remove redundant and unnecessary intermediaries in the patient pathway. It transforms pathology from a results and knowledge service into a transactional service – this may be regarded as the application of Lean methodology.62 The principle is that where a positive test in a care pathway should lead to a specific intervention then this is generated automatically via the laboratory report rather than requiring the specific input of a primary or secondary care clinician. An example in UK practice would be the creation of an appointment for echo-cardiology based on a raised B-type natriuretic peptide result or a colposcopy appointment on the basis of cervical cytology findings. Such interventions may be viewed as disruptive innovation since their introduction causes a radical shift in the care-delivery model.63 Telepathology Telepathology is a sub-specialty of telemedicine, which promises to be of particular use in situations where expertise is located in centres of excellence geographically distant from the site of clinical service delivery. Remote support is not limited to imaging modalities and the definition can be broadened to include any circumstance where remote access enables transfer of expertise, including remotely accessible POCT.64 Digital Imaging Systems The introduction of digital medical image sources in the 1970s and the use of computers in processing these images after their acquisition led the American College of Radiology and the National Electrical Manufacturers Association to form a joint committee in order to create a standard method for the transmission of medical images and their associated information. This standard is referred to as Digital Imaging and Communication in Medicine (DICOM Standard Version 365). A major concern for pathology, where image data is of increasing interest, is that no equivalent data standards exist. Images are being captured and transmitted in many different formats (e.g. bitmap, JPEG, TIFF) with little regard as to how they might be stored, exchanged or analysed in a collaborative manner in future networked systems. Consideration is being given in some centres as to whether images from whatever source (e.g. scanned images, histopathological images, photographs of gels, cytogenetic smears etc), could and should be stored in DICOM Version 3. However, it is still debatable whether this is an appropriate format, e.g. for histopathological images. The best example of the use of DICOM is the Picture



Archiving and Communication Systems (PACS) now in common use in radiology departments throughout the world. Rather than store images in the host pathology system it might be feasible to include them in the local PACS. The adoption of non-standardised image storage systems in pathology raises concern, as incompatibilities in data formats will limit the potential for future exchange of data. Digital pathology (the complete digitisation of pathology slides) allows for a number of potential benefits in the delivery of patient care – slides are easier to access, the risk of slides getting mixed up is radically reduced and patients can receive specialist review more quickly.66 A number of companies are now marketing commercial solutions but uptake is low outside research and teaching facilities.67 There are several reasons for the slow uptake of digital pathology, including lack of regulatory approval, lack of an established cost benefit case and pathologists’ reluctance to use the technology.68 For many of these reasons an underlying cause may be that the currently available digital pathology software is inefficient compared to conventional microscopy and is rejected by users. A multidisciplinary group of researchers at the University of Leeds and Leeds Teaching Hospitals Trust have spent the past four years developing a virtual microscope to make these benefits more widely available. Preparatory work found that it can take 60% longer to make a diagnosis using digital slides due to the level of detail they contain and the size of the images – if they were printed on paper they would be the size of a squash court, yet most manufacturers only provide slow standard PC-based desktop software with small screens.69 The goal of the project has been to create a Virtual Reality microscope that allows for faster diagnoses with the same accuracy as a conventional microscope. The key innovation of the project is to combine expertise in multiple disciplines (pathology, computer graphics, ethnography, psychology and medical imaging) to design a digital pathology system which is truly fit for purpose for both individual pathologists and the health service. The team has developed both a single-user workstation for clinical use, as well as a multi-user ‘Powerwall’ consisting of 12 high-resolution monitors that can be used for teaching and multidisciplinary team meetings.33 Implementation of Call Centre Technology In industry sectors such as banking the use of sophisticated call handling technologies is already well-established and widely used. Such technologies offer many opportunities to link information systems, e.g. by providing a single point of telephone enquiry with call filtering, queuing and internal routing to appropriate services. Potential applications of this technology in pathology include provision and/or coordination of results ‘hot lines’, multidisciplinary advice enquiry points, messaging services, specimen transport,


handbook and protocol supplies, and other activities related to ‘customer service’. An example of such a system was developed in Spokane Washington and subsequently commercialised by Sunquest as Outreach Advantage.70 This used Microsoft Dynamics software to provide a Customer Relationship Management solution which managed services to clients distributed across five Western US States including the management of phlebotomists, sample tracking and complaints management. The benefits were improved customer relations, improved service levels in terms of turnaround times and a growth in the business.71 Managing Laboratories Using Information Systems Pathology information systems will play an increasingly important role in service management, particularly with the expanded scope of pathology networks. Existing systems (e.g. for finance, personnel and resource management) will have to evolve to meet these needs. Modeling of workload and logistics, production of financial forecasts, and detailed cost analysis will be required to maximise efficient delivery of services. As pathology expenditure accounts for approximately 3-10% of health budgets,72 worldwide there is considerable benefit if costs can be controlled. One example where this has already been piloted is on large scale diagnostic instruments, which place automatic orders for reagents on the basis of workload (e.g. Siemens, Roche, Abbott platforms). Traditional LIMS, though good at capturing raw workload data, do not provide the functionality to support ‘back office’ management tasks and most services will employ secondary systems to support this, usually feeding data from the LIMS into a separate database or data warehouse against which standard business reporting tools can be applied. Maintaining the coherence of codes between such systems can be a challenge and usually involves some form of mapping or aggregation function to link laboratory clinical test codes to product codes and price/costing models. LOINC was developed in part to service this function and is comprehensive in coverage but as it was not developed purely for clinical use, it falls short in the clinical domain. This is a generic problem for coding systems which do not transfer well to functions for which they were not designed.73 Resources, Staffing and Skills The ability to update laboratory information systems will depend critically on the availability of resources, both staff time and financial. Though the use of information technology in the laboratory is impressive, the most important success factors are the sophistication of the IT support systems which have been developed and the IT skill base of the staff.


Table 3. Future Informatics skill profiles for pathology staff. Skill Level

Required competencies

Staff to which these apply

Basic

PC literacy including data access and entry, web browsing

Administrative & clerical Technical

Intermediate

General applications, including word processing, spreadsheets, databases

Administrative & clerical Supervisory technical Scientific

Advanced

More advanced applications, including statistical analysis, modelling, bioinformatics, web site design

Scientific Research & development Management

Specialist

Information system strategy, development and other specialist applications, including design of customised management or decision support systems

System managers IT support

This is necessary to create and maintain robust systems of quality assurance of the data handling process and to ensure appropriate use of the systems. An above average level of computer literacy is now a requirement for employment in the laboratory and a current problem is the loss of so many of the best laboratory informatics staff to the hospital IS departments and the IT industry once they have developed their skills. Ideally, all pathology laboratories should have an Information Strategy Group, which has the remit of defining IT requirements to meet the internal and external needs of the service, both current and projected. These groups should be formally represented at the highest levels of management and be treated as a strategic not just technical resource. Laboratories will also need a cohort of suitably qualified IT staff to support the full range of information systems and their exploitation. Ideally, a laboratory systems manager should hold professional qualifications in both laboratory sciences and IT, with some members of the IT team being computer science graduates with expertise in specialist subject areas such as database administration, networking and desktop support. In addition IT literacy amongst all pathology staff will need to be high since systems are integral to all aspects of analytical and managerial activity. All pathology staff will be expected to have basic skills. Achievement of additional competencies, at an appropriate level depending on job function (Table 3), should be anticipated. These would include skills in bioinformatics and statistics for example in areas exploiting the emergent genomic and proteomic technologies. There will be many opportunities for pathologists to develop new roles in informatics support in the future, especially in relation to optimising communications with clinical systems. It is to be

hoped that professional bodies will recognise this and develop the necessary training and accreditation resources. In the US the Association of Pathology Informatics is promoting such training and has links to similar initiatives across many other countries.74 Conclusions The core of any pathology department is its information system. In the last three decades these systems have developed from supporting core pathology services to delivering services throughout the hospital network. The integration of pathology systems and the processes required to support this are becoming ever more complex, reflecting modern requirements for health care delivery, including electronic healthcare records. Pathology data contributes significantly to these records, which will determine the design of the next generation of IT systems. From the discussion so far it should be clear that LIMS systems of the future will be very different animals from those in use today. Current systems are an endangered species whose niche will be supplanted by applications spanning the clinical laboratories interface and providing interactive access to laboratory data and information. The major challenge for laboratory scientists and informaticians is to promote use of their knowledge in the clinical world. As always, the problem is not with the computers but with the people. Optimal use of pathology services will depend on greater involvement of IT in the pre- and post analytical phases, both to manage demand and to encourage appropriate test requesting. This necessitates more sharing of information between different clinical systems, and hence improvements both in standardisation and in communicability. Clinical Clin Biochem Rev 35 (3) 2014 189


governance requires adherence to protocols and guidelines, which can be facilitated by pathology systems. Clinicians are likely to become increasingly dependent on direct access to pathology data. The challenge for laboratory scientists is to ensure the timely and effective introduction of IT systems that can meet the needs of twenty-first century medicine. Acknowledgements We would like to thank our many colleagues in pathology and computer science for their fresh ideas and support. Competing Interests: None declared (RGJ, GB). OAJ is director of X-Lab Systems. References 1. Wootton ID. Computer science in the biochemistry laboratory. Br Med Bull 1968;24:219-23. 2. Whitehead TP, editor. Automation and Data Processing in Pathology. London: College of Pathologists; 1969. 3. Whitehead TP. The computer in the laboratory. Practitioner 1969;203:294-505. 4. Department of Health and Social Security. Annual Review of National Health Service Computing Parts 1 and 2. London: HMSO; 1973. 5. Jones R, Batstone G. Report of an audit by NHS Connecting for Health on test reporting standards in electronic pathology messages sent to primary care. RCPath Bulletin 2012;160:227-30. 6. Miller RE, Causey JP, Moore GW, Wilk GE. Development and Operation of a MUMPS laboratory information system: a decade’s experience. Proc Annu Symp Comput Appl Med Care. 1988:654-8. 7. Independent Review of NHS and Social Care IT, August 2009. http://www.rcplondon.ac.uk/sites/default/files/nhsand-social-care-it-review_0.pdf (Accessed 11 June 2014) 8. Payne TH, Detmer DE, Wyatt JC, Buchan IE. Nationalscale clinical information exchange in the United Kingdom: lessons for the United States. J Am Med Inform Assoc 2011;18:91-8. 9. UK Clinical Research Collaboration. Report of the Research Simulations. 2007. 10. New Centres put health records at the heart of UK medical research. http://www.abdn.ac.uk/news/4440/ (Accessed 1 July 2014). 11. Initiatives in Informatics Research – Medical Research Council. http://www.mrc.ac.uk/research/initiatives/ health-and-biomedical-informatics/initiatives-ininformatics-research/ (Accessed 1 July 2014). 12. Edwards GA. Expert systems for clinical pathology reporting. Clin Biochem Rev 2008;29(Suppl 1):S105-9. 13. Price CP, Jones RG. The challenges in commissioning laboratory medicine (pathology) services. J Manage Market Healthcare 2008;1:166-78. 14. Lundberg GD. Changing physician behavior in ordering diagnostic tests. JAMA 1998;280:2036. 15. Newman G, Jones R. Can the use of different business 190 Clin Biochem Rev 35 (3) 2014

models in transformation and transitioning of back offices for public services be applied to pathology services? RCPath Bulletin 2010;152:294-6. 16. Mechling J. Transforming the Back Office – Why and How? National Association of State Auditors, Comptrollers and Treasurers and the National Electronic Commerce Coordinating Council; 2006. http://www. iccs-isac.org/en/tempdoc/research/Transforming%20 Back%20Office.pdf (Accessed 8 April 2014). 17. Peters M, Broughton PM. The role of expert systems in improving the test requesting patterns of clinicians. Ann Clin Biochem 1993;30:52-9. 18. Wyatt JC. Computer-assisted decision support. In: SosaIudicissa M, Oliveri N, Gamboa CA, Roberts J, editors. Internet, Telematics and Health. Amsterdam, Netherlands: IOS Press; 1997. 19. Croal B. Demand management: controlling laboratory workload. In: Price CP, Christenson RH, editors. Evidence-Based Laboratory Medicine; Principles, Practice and Outcomes. 2nd ed. Washington DC: AACC Press; 2007:399-421. 20. Main C, Moxham T, Wyatt JC, Kay J, Anderson R, Stein K. Computerised decision support systems in order communication for diagnostic, screening or monitoring test ordering: systematic reviews of the effects and cost-effectiveness of systems. Health Technol Assess 2010;14:1-227. 21. Lord Carter of Coles. Report of the Review of NHS Pathology Services in England: An Independent Review for the Department of Health. 2006. 22. Smellie WS, Galloway MJ, Chinn D. Benchmarking general practice use of pathology services: a model for monitoring change. J Clin Pathol 2000;53:476-80. 23. McVittie JD, Kay J, Quainton TJ, Stewart RD. Remote links to a laboratory computer: low-cost solutions without high-speed networking. Ann Clin Biochem 1988;25(S1):187-8. 24. McVittie J, Kay J. Communication of structured medical information by computers: the Oxford experience. In: De Moor GJE, McDonald CJ, Noothoven van Goor J, editors. Progress in Standardisation in Health Care Informatics. Amsterdam: IOS Press; 1993. 25. Tapuria A, Kalra D, Kobayashi S. Contribution of Clinical Archetypes, and the Challenges, towards Achieving Semantic Interoperability for EHRs. Healthc Inform Res 2013;19:286-92. 26. Kalra D, Musen M, Smith B, Ceusters W, De Moor G. ARGOS policy brief on semantic interoperability. Stud Health Technol Inform 2011;170:1-15. 27. Westgard S. Prioritizing risk analysis quality control plans based on Sigma-metrics. Clin Lab Med 2013;33:41-53. 28. Dalenberg DA, Schryver PG, Klee GG. Analytical performance specifications: relating laboratory performance to quality required for intended clinical use. Clin Lab Med 2013;33:55-73. 29. Dr Lal PathLabs Selects STARLIMS Software for LIMS. http://healthcare.financialexpress.com/201111/market18.


shtml (Accessed 1 July 2014). 30. Selby NM, Devonald MAJ. What is the role of e-alerts in acute kidney injury? J R Coll Physicians Edinb 2012; 42(Suppl 19):12-20. 31. Pathology Modernisation Lab to Lab Communication. http://www.ychi.leeds.ac.uk/labtolab (Accessed 20 April 2014). 32. The National Pathology Exchange (NPEx). http://www. npex.nhs.uk (Accessed 20 April 2014). 33. Digital First: Clinical Transformation through Pathology Innovation, NHS National Pathology Programme, Department of Health, London, 2013. 34. Genetics Screening and Testing Services. http://www. pkihs.com/newbornhealth-screening-testing.html (Accessed 20 April 2014). 35. NHS Genetic Service Information Systems Output Based Specification (OBS). Document GEN-3-006. Manchester, UK: National Genetics Reference Laboratory (NGRL); 8 September 2005. 36. Murphy MF, Fraser E, Miles D, Noel S, Staves J, Cripps B, Kay J. How do we monitor hospital transfusion practice using an end-to-end electronic transfusion management system? Transfusion 2012;52:2502-12. 37. Baer D. Patient-physician e-mail communication: the kaiser permanente experience. J Oncol Pract 2011;7:2303. 38. Jones RG, Hawker M, Johnson OA. Informatics in point of care testing. In: Price CP, St. John A, Hicks JM, editors. Point of Care Testing. 3rd ed. Washington, DC: AACC Press; 2010. 39. National service frameworks and strategies. h t t p : / / w w w. n h s . u k / n h s e n g l a n d / N S F / p a g e s / Nationalserviceframeworks.aspx (Accessed 20 April 2014). 40. Health Level Seven (HL7). http://www.hl7.org.uk (Accessed 20 April 2014). 41. Information Standards Board for Health and Social Care. Standard ISB 1077 AIDC for Patient Identification, 2011. http://www.isb.nhs.uk/library/standard/99 (Accessed 20 April 2014). 42. Identifying Patients - Making sure the right patient gets the right treatment. http://www.gs1uk.org/what-we-do/ sector-solutions/healthcare/Pages/Increasing-NHSefficiency-using-GS1-standards.aspx (Accessed 20 April 2014). 43. International Health Terminology Standards Development Organisation (IHTSDO) SNOMED CT. http://www. ihtsdo.org/snomed-ct/ (Accessed 20 April 2014). 44. Benson T. The history of the Read Codes: the inaugural James Read Memorial Lecture 2011. Inform Prim Care 2011;19:173-82. 45. Forrey AW, McDonald CJ, DeMoor G, Huff SM, Leavelle D, Leland D, et al. Logical observation identifier names and codes (LOINC) database: a public use set of codes and names for electronic reporting of clinical laboratory test results. Clin Chem 1996;42:81-90. 46. Jones R, Batstone G, Gutteridge C, Croal B, Barnes I.

NLMC-more than just for pathology. RCPath Bulletin 2012;160:233-6. 47. Pathology Units and Terminology Standardisation Project. http://www.rcpa.edu.au/Library/PractisingPathology/PTIS/PUTS (Accessed 3 June 2014). 48. Lundberg GD. The need for an outcomes research agenda for clinical laboratory testing. JAMA 1998;280:565-6. 49. Price CP, Christenson RH, eds. Evidence-Based Laboratory Medicine: Principles, Practice and Outcomes. 2nd ed. Washington DC: AACC Press; 2007. pp. 1-545. 50. Price CP. Evidence-based laboratory medicine: is it working in practice? Clin Biochem Rev 2012;33:13-9. 51. Driskell OJ, Holland D, Hanna FW, Jones PW, Pemberton RJ, Tran M, et al. Inappropriate requesting of glycated hemoglobin (Hb A1c) is widespread: assessment of prevalence, impact of national guidance, and practice-topractice variability. Clin Chem 2012;58:906-15. 52. Welsh Clinical Portal. http://www.wales.nhs.uk/nwis/ page/52547 (Accessed 20 April 2014). 53. UK Standards for Microbiology Investigations (SMI). http://www.hpa.org.uk/SMI (Accessed 20 April 2014). 54. Reynolds TM, Penney MD. The mathematical basis of multivariate risk screening: with special reference to screening for Down’s syndrome associated pregnancy. Ann Clin Biochem 1990;27:452-8. 55. Reynolds T, Ellis A, Jones R. Down’s syndrome risk estimates demonstrate considerable heterogeneity despite homogeneity of input. Ann Clin Biochem 2004;41:464-8. 56. Dati F. The new European directive on in vitro diagnostics. Clin Chem Lab Med 2003;41:1289-98. 57. Raebel MA, Haynes K, Woodworth TS, Saylor G, Cavagnaro E, Coughlin KO, et al. Electronic clinical laboratory test results data tables: lessons from MiniSentinel. Pharmacoepidemiol Drug Saf 2014;23:609-18. doi:10.1002/pds.3580. 58. Assay Finder. http://www.assayfinder.com/ (Accessed 20 April 2014). 59. National Institute for Health and Care Excellence (NICE). https://www.evidence.nhs.uk/ (Accessed 20 April 2014). 60. Komatireddy R, Topol EJ. Medicine unplugged: the future of laboratory medicine. Clin Chem 2012;58:16447. 61. Eysenbach G. From intermediation to disintermediation and apomediation: new models for consumers to access and assess the credibility of health information in the age of Web2.0. Stud Health Technol Inform 2007;129:162-6. 62. Krafcik JF. Triumph of the lean production system. Sloan Manage Rev 1988;30:41-52. 63. Hwang J. Keynote address—the innovator’s prescription: an examination of the future of health care through the lenses of disruptive innovation. Arch Pathol Lab Med 2009;133:513-20. 64. Weinstein RS, Graham AR, Richter LC, Barker GP, Krupinski EA, Lopez AM, et al. Overview of telepathology, virtual microscopy, and whole slide imaging: prospects for the future. Hum Pathol 2009;40:1057-69. 65. Digital Imaging and Communications in Medicine Clin Biochem Rev 35 (3) 2014 191


(DICOM). http://dicom.nema.org/ (Accessed 20 April 2014). 66. Park S, Parwani AV, Aller RD, Banach L, Becich MJ, Borkenfeld S, et al. The history of pathology informatics: A global perspective. J Pathol Inform 2013;4:7. 67. Rojo MG, García GB, Mateos CP, García JG, Vicente MC. Critical comparison of 31 commercially available digital slide systems in pathology. Int J Surg Pathol 2006;14:285-305. 68. Pantanowitz L, Wiley CA, Demetris A, Lesniak A, Ahmed I, Cable W, et al. Experience with multimodality telepathology at the University of Pittsburgh Medical Center. J Pathol Inform 2012;3:45. 69. Randell R, Ruddle RA, Mello-Thoms C, Thomas RG, Quirke P, Treanor D. Virtual reality microscope versus conventional microscope regarding time to diagnosis: an experimental study. Histopathology 2013;62:351-8. 70. Outreach Advantage Customer Relationship Management - Sunquest Information Systems. http:// www.sunquestinfo.com/products-solutions/outreach/oacustomer-relationship-management/ (Accessed 20 April 2014). 71. Johnston M. The impact of information technology advances on laboratory medicine. Frontiers in Laboratory Medicine 29-30 January 2008, Manchester, UK. 72. International Profiles of Health Care Systems. The Commonwealth Fund, June 2010 http://www. commonwealthfund.org/~/media/Files/Publications/ Fund%20Report/2010/Jun/1417_Squires_Intl_ Profiles_622.pdf (Accessed 20 April 2014). 73. Coiera E. Guide to Medical Informatics, the Internet and Telemedicine. 2nd ed. London: Chapman & Hall Medical; 2003 74. Association for Pathology Informatics. Training Opportunities in Pathology Informatics. http://www. pathologyinformatics.org/content/training-opportunitiespathology-informatics (Accessed 3 June 2014).


Obituary

Richard Gwent Jones, a champion of health informatics and laboratory medicine, died on 1st July 2014, aged 62. Rick Jones may not be known to many in the Australian laboratory community but for those of us with an interest in IT and informatics, he was regarded both as a visionary and as someone who well understood the particular challenges associated with IT implementation. His knowledge and understanding of this complex area was immense and he was incredibly generous with his advice and help. The fact that many sought his assistance was in part because he had that rare ability in IT of being able to communicate effectively with the most technophobic of individuals! After Rick completed his medical degree and subsequent diabetes research at Oxford, he foresaw that a career in Chemical Pathology might foster his abiding interest in the role of computers in medicine, and he moved to Leeds where he spent the remainder of his career. He had both academic and service appointments within individual hospitals and then across what was, at the time, the largest pathology service in the country - where he had the important strategic role of ensuring consistency of service and results across the whole network. Rick introduced many innovations in his time at Leeds and was instrumental in the establishment of the Yorkshire Centre for Health Informatics and, more recently, the MRC Centre for Bioinformatics. Rick saw the power of data long before ‘big data’ became the much discussed topic it is now and he established systems in the UK to interrogate huge NHS databases of pathology results to provide immensely valuable information, a resource that we can only envy here in Australia given the fragmented nature of our own pathology system. Rick visited Australia several times, first to the seminal Informatics meeting at Uluru in 1993 and later he talked at The Business of Pathology meeting in Sydney in 2007. Some of us would have had him visit here more often but for his reluctance to travel long distances following his treatment for a haematological cancer. His contribution to laboratory IT, particularly in the UK has been immense, and this was recognised earlier this year when he was awarded a personal Chair in Health Informatics and Chemical Pathology. Rick agreed to write the Informatics review in this issue of The Clinical Biochemist Reviews when he was relatively well and still working. Early this year he learnt that there were no more treatment options left and retired to home. Despite that he delivered the review and responded to reviewers’ comments just a few weeks before he passed away. Another measure of his commitment and generosity. Andrew St John


Good laboratory practice for clinical next-generation sequencing informatics pipelines.

The potential impact of laboratory informatics on technology transfer.

Clinical microbiology informatics.

Optimizing Clinical Research Participant Selection with Informatics.

Guest-editorial: Biomedical informatics in clinical environments.

Medical student awareness of and interest in clinical informatics.

Melioidosis and safety in the clinical laboratory.

Biobanking informatics infrastructure to support clinical and translational research.

A review of analytics and clinical informatics in health care.

Personalized medicine and the clinical laboratory.

Clinical Research Informatics: Recent Advances and Future Directions.

Clinical research informatics and electronic health record data.

The Clinical and Health Economic Value of Clinical Laboratory Diagnostics.

Perspectives on clinical informatics: integrating large-scale clinical, genomic, and health information for clinical care.

Amebiasis: clinical and laboratory perspectives.

Building blocks for a clinical imaging informatics environment.

Clinical laboratory directors.

Laboratory informatics based evaluation of methylene tetrahydrofolate reductase C677T genetic test overutilization.

Milestones: Critical Elements in Clinical Informatics Fellowship Programs.

Molecular Technologies in the Clinical Diagnostic Laboratory.

Clinical proteome informatics workbench detects pathogenic mutations in hereditary amyloidoses.

Clinical informatics: prospects for a new medical subspecialty.

Risk management in the clinical laboratory.

[Quality control in the clinical biology laboratory].