Full article title Diagnostic informatics: The role of digital health in diagnostic stewardship and the achievement of excellence, safety, and value
Journal Frontiers in Digital Health
Author(s) Georgiou, Andrew; Li, Julie; Hardie, Rae-Anne; Wabe, Nasir; Horvath, Andrea R.; Post, Jeffrey J.; Eigenstetter, Alex; Lindeman, Robert; Lam, Que; Badrick, Tony; Pearce, Christopher
Author affiliation(s) Macquarie University, Prince of Wales Hospital, University of New South Wales, New South Wales Health Pathology, Austin Health, Royal College of Pathologists of Australasia, Outcome Health, Monash University
Primary contact Email: andrew dot georgiou at mq dot edu dot au
Editors Chapman, Wendy
Year published 2021
Volume and issue 3
Article # 659652
DOI 10.3389/fdgth.2021.659652
ISSN 2673-253X
Distribution license Creative Commons Attribution 4.0 International
Website https://www.frontiersin.org/articles/10.3389/fdgth.2021.659652/full
Download https://www.frontiersin.org/articles/10.3389/fdgth.2021.659652/pdf (PDF)

Abstract

Diagnostic investigations (i.e., pathology laboratory analysis and medical imaging) aim to increase the certainty of the presence of or absence of disease by supporting the process of differential diagnosis, support clinical management, and monitor a patient's trajectory (e.g., disease progression or response to treatment). Digital health—defined as the collection, storage, retrieval, transmission, and utilization of data, information, and knowledge to support healthcare—has become an essential component of the diagnostic investigational process, helping to facilitate the accuracy and timeliness of information transfer and enhance the effectiveness of decision-making processes. Digital health is also important to diagnostic stewardship, which involves coordinated guidance and interventions to ensure the appropriate utilization of diagnostic tests for therapeutic decision-making. Diagnostic stewardship and informatics are thus important in efforts to establish shared decision-making. This is because they contribute to the establishment of shared information platforms (which enable patients to read, comment on, and share in decisions about their care) based on timely and meaningful communication.

This paper will outline key diagnostic informatics and stewardship initiatives across three interrelated fields: (1) diagnostic error and the establishment of outcome-based diagnostic research, (2) the safety and effectiveness of test result management and follow-up, and (3) digitally enhanced clinical decision support systems.

Keywords: pathology, diagnostic error, outcomes based assessment, evaluation, safety, health informatics, decision support, test result follow-up

Introduction

Diagnostic investigations and tests (e.g., pathology and medical imaging) involve the observation of personal characteristics, symptoms, signs, and history. Their aim is to increase certainty of the presence of or absence of disease (or in the case of the differential diagnosis process, distinguish between different possible diagnoses), support clinical management, and monitor a patient's trajectory (e.g., during or after treatment).[1] Most recently, diagnostic investigations have had a pivotal role in the response to the SARS-CoV-2 pandemic through laboratory testing and epidemiological surveillance of viral infection, not only from its detection and diagnosis (e.g., via molecular testing by real-time polymerase chain reaction [RT-PCR]) but also its prognostication, treatment, monitoring (e.g., testing related to comorbidities), and recovery (aided by anti-SARS-CoV-2 monoclonal antibodies).

Digital health—defined as the collection, storage, retrieval, transmission, and utilization of data, information, and knowledge to support healthcare[2]—contributes to effective decision-making during diagnostic investigations by improving the ability to gather, organize, and display information, or by enhancing timely access to diagnostic reference information, as a result facilitating follow-up and providing valuable feedback to clinicians and patients.[3] Diagnostic informatics—the use of digital health tools to facilitate the accuracy and timeliness of health information transfer and enhance the effectiveness of the decision-making processes[4]—is fundamental for the safe and effective management of diagnostic investigations and the data that comes from them. Diagnostic informatics is thusly a key component of successful diagnostic stewardship, which encompasses the coordination of guidance and interventions to ensure the appropriate utilization of diagnostic tests for therapeutic decision-making.[5]

This paper will outline the importance of key diagnostic informatics and digital health stewardship initiatives across three interrelated areas: (1) diagnostic error and the establishment of outcome-based diagnostic research; (2) the safety and effectiveness of test result management and follow-up, including the critical role that patients can play in the process; and (3) digitally enhanced clinical decision support systems.

Diagnostic error and the establishment of outcome-based diagnostic informatics research

Diagnostic error is a major problem in healthcare, contributing to ~10% of patient deaths and between 6 to 17% of hospital adverse events.[6] Diagnostic error involves either the failure to establish an accurate and timely explanation of the patient's health problem(s) or the failure to effectively communicate that explanation to the patient.[6] In Australia, medico-legal data show that diagnostic error is implicated in half of medical negligence claims involving general practitioners.[7]

Many factors can contribute to diagnostic error, including the lack of an integrated care pathway; the involvement of multiple specialists; problems with collaboration and communication among clinicians, patients, and their families; lack of infrastructure to support the diagnostic process; and inadequate attention to understanding the health problem and its causes.[6] Existing evidence shows that in acute care, an estimated 45% of laboratory testing is underutilized (i.e., when one or more tests should have been undertaken but weren't) and 21% is overutilized (i.e., when one or more tests were unnecessary or repeated within an inappropriate time frame).[8]

There is currently a lack of a systematic outcomes-based approach to identify and monitor the prevalence and impact of diagnostic errors. In part, this is because the typical relationship between a test and a health outcome is indirect. It is made harder still by the existence of data silos across different clinical and care settings (e.g., pathology, medical imaging, medical records, emergency, and hospital administration systems), which limit the ability to link test results and referrals to the different components of the patient journey (e.g., treatment and outcome). Despite the existence of evidence-based guidelines to encourage best practice utilization of diagnostic tests, there are very few studies on how these guidelines have impacted testing patterns and patient outcomes. This was noted by a recent international scoping review that drew attention to the significant lack of strategies for optimizing test utilization and improving patient outcomes in general practice.[9]

Linked and integrated digital health data sources and repositories across hospital, pathology, and primary care provide a means to identify, measure, and monitor the quality of care. They can deliver key outcome-based measures of diagnostic utilization such as patient outcome or hospital admissions, which can be used to evaluate the impact of vital interventions (e.g., electronic decision support). This is the premise underlying critical initiatives like the New South Wales (NSW) Health Pathology Atlas of Variation (in collaboration with the NSW Emergency Care Institute and NSW eHealth Integrated Care), designed to create a statewide quality improvement project.[10]

As an example of the potential of outcome-focused studies to impact on patient care, we investigated the incidence of acute kidney injury (AKI) over a five-year period (2009–2013) across four NSW hospitals.[11] Patients with AKI were identified using the serum creatinine-based definition as described in the international consensus guidelines published by the “Kidney Disease: Improving Global Outcomes” (KDIGO) Work Group. This required programming of a complex algorithm, which was run on more than two million creatinine results to detect the presence of AKI. The study found that only 15.9% of hospitalizations with AKI stage 1 were coded as such[11], thus highlighting the need for an evidence-based laboratory AKI decision support aid. Other examples of the value of linked data helping to power programs of research include measuring the effects of electronic medical records (EMRs) on healthcare[12][13], and the impact of the rapid flu tests on emergency department (ED) performance and patient outcomes.[14]

Enhancing the safety and effectiveness of test result management and follow-up

A significant source of diagnostic error can be attributed to shortcomings in the follow-up of test results, identified as a priority area by the World Health Organization's World Alliance for Patient Safety.[15] International evidence shows between 41 and 100% of patients leave the hospital with at least one test result pending at discharge, influenced in part by pressure on doctors to reduce length of stay.[16]

Care transitions can be defined as the passage of patients among healthcare professionals, settings (e.g., units, wards, hospitals, aged care facilities), and their home.[17] Ineffective care transitions can lead to adverse events, such as medication errors[18], medical errors related to the completion of diagnostic workups[19], and the loss of crucial patient information.[17] Arriving at a diagnosis is not a single task but rather a series of tasks and responsibilities over numerous time points, shared among many people across the healthcare spectrum.[20]

Pathology and medical imaging services perform a major role in patient care by ensuring that reliable and accurate results are delivered in a timely fashion to inform clinical management. Nevertheless, there are large variations in how critically abnormal results are defined and reported, and the failure of laboratories to uniformly follow internationally recognized guidelines.[21] Several initiatives seek to address these issues, including The Royal College of Pathologists of Australasia (RCPA)—Australasian Association of Clinical Biochemists (AACB) High Risk Results Working Party, which is developing a harmonized, evidence-based “Alert List for Australasian Laboratories.”[22] Similarly, the RCPA-led Pathology Information, Terminology and Units Standardizations (PITUS) program aims to establish standardized pathology information structures and terminologies to improve recording, decision support, and communication of laboratory information.[23][24]

Another initiaitve worth noting is a national Australian stakeholder forum (involving patient groups, clinicians, managers, and professional organizations), held in Sydney in 2016 to identify patient safety challenges related to shortcomings in the governance and integration of the test result management process and the underutilization of digital health solutions. The forum drew particular attention to the role patients can contribute to effective test result management. It recommended robust follow-up procedures particularly for tests that may be pending during transitions of care, critical test results, and non-critical but actionable test results, which can have a severe impact on patients' health outcomes.

Drawing from these and other such examples, diagnostic informatics, along with diagnostic stewardship, provide a foundation stone for improved patient engagement and shared decision-making. This is because they incorporate shared information (enabling patients to read, comment on, and share in decisions about their care) and timely and meaningful communication (enabling patients to receive, send, and comprehend the information required).[25] By supporting patients to make informed decisions about their own health and the care they receive, this approach promotes person-centered care, an established key driver of quality in healthcare[26] (as depicted in Figure 1).


Fig1 Georgiou FrontDigHlth2021 3.jpg

Figure 1. Key features of person-centered care.

Digitally enabled and enhanced decision support

Digital health can contribute to effective decision-making by improving the ability to gather, organize, and display information, or by enhancing timely access to diagnostic reference information, facilitating follow-up and feedback to clinicians and patients. Electronic support can be used to[27]:

  • optimize decision-making (e.g., choice of an appropriate laboratory test, improved compliance with care guidelines, improved interpretation of test results);
  • improve care processes (e.g., better documentation and improved communication); and
  • prevent errors of commission (e.g., unnecessary repetition of laboratory tests) and errors of omission (e.g., trigger immediate action in response to high-risk laboratory results).

Digitally enabled decision support tools can be used in a wide variety of diagnostic-related areas, including as alerts and order sets (to enhance the appropriate choice of tests), as a provision of important information related to decision-making (to support the identification and management of risk factors), as guideline support, and as a provision of relevant reference material (to promote the best available care).

The Sensible Test Ordering Practice (STOP) initiative was developed in Australia to promote consistent and rational diagnostic test ordering practices in acute care settings. STOP involves the use of a traffic-light system (green, amber, or red) to restrict the range of tests that can be ordered depending on the seniority of the clinicians: “green” tests may be ordered by junior doctors, while more complex, expensive, or lower-yield tests designated “amber” or “red” require senior medical staff sign off. The Australasian College for Emergency Medicine and the RCPA have developed a set of laboratory testing recommendations in EDs for common presentations based on the STOP principles.[28] An example for adults presenting to EDs with chest pain is presented in Table 1. The STOP guidelines can be integrated into electronic decision support tools linked to EMR systems to facilitate the consistency and appropriateness of diagnostic-test-ordering practices.


Tab1 Georgiou FrontDigHlth2021 3.jpg

Table 1. STOP principles for adults presenting to ED with chest pain

While electronic decision support systems are often described as key to reducing misdiagnosis, their impact has not been strongly validated against patient outcomes, and their diffusion is patchy.[29] There is also a long history of failure associated with the use of computer-based tools.[29] These tools are often introduced to a clinical setting with limited pilot testing and little understanding of how they affect workflow.[29] Establishing effective digital health interventions must ensure the acceptance and usability of the digital health system and include systematic feedback about their performance to users.[3][30] The STOP electronic decision support initiative will thus need to be evaluated to ensure that it is effective (e.g., helps to improve the quality of care); efficient (e.g., is accessible and responsive to clinician and patient needs); appropriate (e.g., promotes the best possible outcomes of care); and safe (e.g., helps to optimize care quality and safety).

Discussion

The diagnostic process occurs over time and often involves multiple healthcare professionals across different care settings. Arriving at a diagnosis can be an uncertain process, often reliant on limited and imperfect data or indeterminate diagnostic interpretations from pathology reports.[6][31] The measurement of diagnosis is further challenged by traditional health record structures, which can fail to capture the crucial clinician and patient narrative, communication context, and organizational culture.[31] Recent advances in artificial intelligence and natural language processing tools (such as machine translation, speech recognition, and speech synthesis) provide fertile ground to improve and enhance our understanding of the diagnostic process.[32]

To ensure the seamless flow of, and access to, patient information, interoperable electronic health systems are critical.[6] However, a systematic review of studies of the impact of those digital health systems on test result management has shown that, in and of itself, digital health does not provide a complete solution.[33] The development of safe and effective digital health interventions needs to encompass multiple dimensions, taking into account patient-centered care (involving shared, timely, and meaningful information); existing diagnostic processes (involving multiple people across different clinical settings); and key organizational communication processes (multiple transactions requiring feedback, iteration, and confirmation).[33][34][35]

Acknowledgements

Author contributions

All authors contributed to the conception, design, and drafting of the work, drafting of the work, revising it critically for important intellectual content; provided final approval of the version to be published; and provided agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

The research described in this article was funded by a National Health and Medical Research Council Partnership Grant (APP1111925) involving Macquarie University, NSW Health Pathology, and the Australian Commission on Safety and Quality in Health Care.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

  1. Knottnerus, J.A.; van Weel, C.; Muris, J.W.M. (2002). "Evaluation of diagnostic procedures". BMJ 324 (7335): 477-80. doi:10.1136/bmj.324.7335.477. PMC PMC1122397. PMID 11859054. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1122397. 
  2. Rowlands, D. (2 December 2019). "What is digital health? And why does it matter?" (PDF). Digital Health Workforce Academy. https://www.hisa.org.au/wp-content/uploads/2019/12/What_is_Digital_Health.pdf?x97063. 
  3. 3.0 3.1 El-Kareh, R.; Hasan, O.; Schiff, G.D. (2013). "Use of health information technology to reduce diagnostic errors". BMJ Quality and Safety 22 (Suppl 2): ii40-ii51. doi:10.1136/bmjqs-2013-001884. PMC PMC3786650. PMID 23852973. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3786650. 
  4. Aller, R.D.; Georgiou, A.; Pantanowitz, L. (2012). "Electronic Health Records". In Pantanowitz, L.; Tuthill, J.M.; Balis, U.. Pathology Informatics: Theory and Practice. pp. 217–30. ISBN 9780891896197. 
  5. Global Antimicrobial Resistance Surveillance System (2016). "Diagnostic stewardship: A guide to implementation in antimicrobial resistance surveillance sites". World Health Organization. https://www.who.int/publications/i/item/WHO-DGO-AMR-2016.3. 
  6. 6.0 6.1 6.2 6.3 6.4 Balogh, E.P.; Miller, B.T.; Ball, J.R., ed. (2015). Improving Diagnosis in Health Care. The National Academies Press. doi:10.17226/21794. ISBN 9780309377706. 
  7. Bird, S. (23 February 2018). "Reducing diagnostic error". newsGP. https://www1.racgp.org.au/newsgp/professional/reducing-diagnostic-error. Retrieved 18 January 2019. 
  8. Zhi, M.; Ding, E.L.; Theisen-Toupal, J. et al. (2013). "The landscape of inappropriate laboratory testing: A 15-year meta-analysis". PLoS One 8 (11): e78962. doi:10.1371/journal.pone.0078962. PMC PMC3829815. PMID 24260139. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3829815. 
  9. Elwenspoek, M.M.C.; Scott, L.J.; Alsop, K. et al. (2020). "What methods are being used to create an evidence base on the use of laboratory tests to monitor long-term conditions in primary care? A scoping review". Family Practice 37 (6): 845-853. doi:10.1093/fampra/cmaa074. PMC PMC7759753. PMID 32820328. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7759753. 
  10. Scowen, C.; Wabe, N.; Eigenstetter, A. et al. (2020). "Evaluating the long-term effects of a data-driven approach to reduce variation in emergency department pathology investigations: study protocol for evaluation of the NSW Health Pathology Atlas of variation". BMJ Open 10 (10): e039437. doi:10.1136/bmjopen-2020-039437. PMC PMC7552857. PMID 33046472. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7552857. 
  11. 11.0 11.1 Campbell, C.A.; Ling, L.; Kotwal, S. et al. (2020). "Under-detection of acute kidney injury in hospitalised patients: A retrospective, multi-site, longitudinal study". Internal Medicine Journal 50 (3): 307-314. doi:10.1111/imj.14264. PMID 30816607. 
  12. Georgiou, A.; Vecellio, E.; Toouli, G. et al. (2013). "Monitoring the impact of the electronic medical record on the quality of laboratory test ordering practices". Studies in Health Technologies and Informatics 188: 33–8. PMID 23823285. 
  13. Vecellio, E.; Li, L.; Xiong, J. et al. (2015) (PDF). Examination of variation in hospital pathology investigations by Diagnosis-Related Groups and associations with outcomes and costs. Centre for Health Systems and Safety Research. ISBN 9781741384307. https://www1.health.gov.au/internet/main/publishing.nsf/Content/0E3978777CE9637BCA2580C200088F78/$File/Final%20Report%20-%20Variation%20in%20Hospital%20Pathology%20-%20Macquarie%20University.pdf. 
  14. Wabe, N.; Li, L.; Lindeman, R. et al. (2019). "The impact of rapid molecular diagnostic testing for respiratory viruses on outcomes for emergency department patients". Medical Journal of Australia 210 (7): 316–20. doi:10.5694/mja2.50049. PMC PMC6617970. PMID 30838671. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6617970. 
  15. World Alliance for Patient Safety (2008). Jha, A.. ed. Summary of the evidence on patient safety: Implications for research. World Health Organization. ISBN 9789241596541. https://apps.who.int/iris/handle/10665/43874. 
  16. Callen, J.; Georgiou, A.; Li, J. et al. (2011). "The safety implications of missed test results for hospitalised patients: A systematic review". BMJ Quality and Safety 20 (2): 194-9. doi:10.1136/bmjqs.2010.044339. PMC PMC3038104. PMID 21300992. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3038104. 
  17. 17.0 17.1 National Quality Forum (September 2017). "Improving Diagnostic Quality and Safety Final Report". National Quality Forum. https://www.qualityforum.org/Publications/2017/09/Improving_Diagnostic_Quality_and_Safety_Final_Report.aspx. Retrieved 02 May 2018. 
  18. Kripalani, S.; Jackson, A.T.; Schnipper, J.L et al. (2007). "Promoting effective transitions of care at hospital discharge: A review of key issues for hospitalists". Journal of Hospital Medicine 2 (5): 314-23. doi:10.1002/jhm.228. PMID 17935242. 
  19. Moore, C.; Wisnivesky, J.; Williams, S. et al. (2003). "Medical errors related to discontinuity of care from an inpatient to an outpatient setting". Journal of General Internal Medicine 18 (8): 646-51. doi:10.1046/j.1525-1497.2003.20722.x. PMC PMC1494907. PMID 12911647. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1494907. 
  20. Carayon, P.; Karsh, B.T.; Cartmill, R. et al. (2010). "Incorporating health information technology into workflow redesign: Request for information summary report". Agency for Healthcare Research and Quality. https://digital.ahrq.gov/ahrq-funded-projects/incorporating-health-information-technology-workflow-redesign/citation/workflow. 
  21. Campbell, C.; Horvath, A. (2012). "Towards harmonisation of critical laboratory result management - Review of the literature and survey of australasian practices". Clinical Biochemist Reviews 33 (4): 149–60. PMC PMC3529552. PMID 23267247. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3529552. 
  22. Campbell, C.; Georgiou, A.; Westbrook, J.I. et al. (2016). "What Alert Thresholds Should Be Used to Identify Critical Risk Results: A Systematic Review of the Evidence". Clinical Chemistry 62 (11): 1445–57. doi:10.1373/clinchem.2016.260638. PMID 27565161. 
  23. Pathology Information, Terminology and Units Standardisation Program (July 2013). "PITUS Update - Newsletter of the RCPA Pathology Information, Terminology and Units Standardisation Program" (PDF). PITUS Update (Royal College of Pathologists of Australasia) (1). https://www.rcpa.edu.au/Library/Practising-Pathology/PTIS/PITUS/Docs/PITUS-Update-Issue-1-July-2013. 
  24. Hardie, R.-A.; Moore, D.; Holzhauser, D. et al. (2018). "Informatics External Quality Assurance (IEQA) Down Under: evaluation of a pilot implementation". Journal of Laboratory Medicine 42 (6): 297-304. doi:10.1515/labmed-2018-0050. 
  25. Joosten, E.A.G.; DeFuentes-Merillas, L.; de Weert, G.H. et al. (2008). "Systematic review of the effects of shared decision-making on patient satisfaction, treatment adherence and health status". Psychotherapy and Psychosomatics 77 (4): 219–26. doi:10.1159/000126073. PMID 18418028. 
  26. de Silva, D. (March 2014). "Helping measure person-centred care: A review of evidence about commonly used approaches and tools used to help measure person-centred care". The Health Foundation. https://www.health.org.uk/publications/helping-measure-person-centred-care. 
  27. Osheroff, J.; Pifer, E.; Teich, J. et al. (2005). Improving Outcomes with Clinical Decision Support: An Implementer's Guide (1st ed.). Productivity Press. ISBN 9780976127727. 
  28. Australasian College for Emergency Medicine, Royal College of Pathologists of Australasia (December 2018). "Pathology Testing in the Emergency Department" (PDF). Royal College of Pathologists of Australasia. https://www.rcpa.edu.au/Library/College-Policies/Guidelines/Pathology-Testing-in-the-Emergency-Department.aspx. 
  29. 29.0 29.1 29.2 Garg, A.X.; Adhikari, N.K.J.; McDonald, H. et al. (2005). "Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: A systematic review". JAMA 293 (10): 1223-38. doi:10.1001/jama.293.10.1223. PMID 15755945. 
  30. Georgiou, A.; Prgomet, M. (2019). "E-health: Developing trust, confidence, quality and sustainability". Health Information Management 48 (2): 59–61. doi:10.1177/1833358318808064. PMID 30392402. 
  31. 31.0 31.1 Miao, M.; Dahm, M.R.; Li, J. et al. (2020). "Managing Uncertainty During the Communication of Diagnostic Test Information Between Patients and Clinicians in Australian Emergency Care". Qualitative Health Research 30 (8): 1287-1300. doi:10.1177/1049732320913037. PMID 32249721. 
  32. Hirschberg, J.; Manning, C.D. (2015). "Advances in natural language processing". Science 349 (6245): 261–6. doi:10.1126/science.aaa8685. PMID 26185244. 
  33. 33.0 33.1 Georgiou, A.; Li, J.; Thomas, J. et al. (2019). "The impact of health information technology on the management and follow-up of test results - A systematic review". JAMIA 26 (7): 678–688. doi:10.1093/jamia/ocz032. PMC PMC6562156. PMID 31192362. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6562156. 
  34. Georgiou, A.; Westbrook, J.I.; Braithwaite, J. (2011). "Time matters: A theoretical and empirical examination of the temporal landscape of a hospital pathology service and the impact of e-health". Social Science and Medicine 72 (10): 1603-10. doi:10.1016/j.socscimed.2011.03.020. PMID 21497430. 
  35. Georgiou, A.; Westbrook, J.I.; Braithwaite, J. (2012). "An empirically-derived approach for investigating Health Information Technology: The Elementally Entangled Organisational Communication (EEOC) framework". BMC Medical Informatics and Decision Making 12: 68. doi:10.1186/1472-6947-12-68. PMC PMC3407796. PMID 22788698. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3407796. 

Notes

This presentation is faithful to the original, with only a few minor changes to presentation, though grammar and word usage was substantially updated for improved readability. In some cases important information was missing from the references, and that information was added.