Health Information Futures

From SystemsWiki

Jump to: navigation, search

Contents

Learning Objectives

In this section the goal is to try and alert managers to technological and Health IT trends which are likely to make a major difference and which at present are not yet fully developed but are showing distinct promise of making a difference.

Inevitably the areas discussed will be somewhat speculative as the future is hardly certain both with respect to technological discoveries and health sector discoveries. Also, of course, the choice of areas covered is rather eclectic rather than exhaustive.

Broadly the objective is to have readers come away from this section with an awareness that what is the norm now is very unlikely to be persistent over even the medium term let alone the long.

Personalised Health

It turns out that the old adage that ‘every-one is different’ is a very important observation. An inherent limitation with much treatment offered by medicine is that the treatment is based on groups of patients - identified by a single diagnosis - who may, and more often are not, a homogeneous group. Increasingly it is becoming clear that there are many apparently single diagnoses which do not capture the diversity and differences of the patients that may be contained within the group.

In the last few years it has become very obvious that, in the treatment of many serious diseases, there are a very wide range of levels of responsiveness to apparently identical treatments given for what, on the surface, seems to be the same disease.

The following describes three major disease areas where change has already happened.

Current applications of personalized medicine

While it may be decades before we see the full benefits of personalized medicine, initial benefits are already here. For example, genetic analysis of patients dealing with blood clots, colorectal cancer, and breast cancer are driving treatment advantages that, until recently, were impossible:

Blood clots. Before the availability of genome-based molecular screening, the dosing of Warfarin, which is prescribed 21 million times a year, was a dangerous game in which too little of the drug could trigger more clots and too much could lead to excessive bleeding.

Since 2007, the U.S. Food and Drug Administration has recommended genotyping for all patients being assessed for therapy involving Warfarin. Genotyping allows prescription of drug therapy regimens only to individuals expected to benefit from that specific drug at that specific dosage.

Colorectal cancer. Metastatic colorectal cancer kills 50,000 Americans every year, more lives than are lost to breast cancer and AIDS combined. Among the drugs most frequently used in treating colon cancer is cetuximab (sold as Erbitux by Bristol-Myers Squibb).

For colon cancer patients, the biomarker that predicts how a tumor will respond to certain drugs is a protein encoded by the KRAS gene, which can be now be determined through a simple test. Because cetuximab is effective only in colon cancer patients with normal KRAS protein, treatment with the drug can be withheld from the 40 percent of patients for whom it would prove ineffective. Alternative therapies can be pursued immediately instead.

Breast Cancer. Just as molecular diagnostic testing of tumors determines which colon cancer patients are most likely to benefit from drug therapy using cetuximab, women with breast tumors can be screened to determine which receptors, if any, their tumor cells contain.

For example, the cells of the highly aggressive "triple-negative" breast cancer have no estrogen, progesterone, or human epidermal growth factor receptors, which are essential to the efficacy of current anti-breast cancer therapies. The application of personalized medicine eliminates both the considerable expense and precious time of trial-and-error treatments and helps clinicians to determine quickly which breast cancer therapies are most likely to succeed.

Source is here:

http://genetichealth.jax.org/personalized-medicine/what-is/applications.html

The site from which it is drawn offers a pretty rich view of where the possibilities lie and the range of illnesses that may potentially have their management modified.

The link between technology and personalised medicine is the technological computational grunt is needed to assess the genomes of the numbers of different patients of the wide range of diseases.

This work is clearly one that is only beginning. One way it is being presently being explored is to take patient disease and treatment information (from their electronic health records) and compare their genomes to try and identify what relationships there are that might then help categorise treatment responders and treatment non responders. The computational effort in doing the genetic sequencing and the information matching is very large indeed.

It is of note that the first sequence of the Human Genome cost billions of dollars to create and that just 13 years later the cost is getting down to the order of $1,000. Reports of such work appeared in the New York Times recently.

See here:

Linking Genes to Diseases by Sifting Through Electronic Medical Records

By CARL ZIMMER

The days of scrawled doctor’s notes are slowly coming to a close. In the United States, 93 percent of hospitals are now using at least some electronic medical records and 2.2 percent have given up paper records completely, according to the consulting firm HIMSS Analytics.

The federal government has been pushing for electronic medical records for a decade, arguing that they will improve health care and bring down costs. That is still a matter of debate. Critics charge that the system is hobbled by poorly designed software and that some hospitals are using electronic medical records to bill more for the same services.

But a new study suggests that electronic medical records may have another, entirely different use: as a Rosetta Stone for our DNA. Researchers are using them to trace links between genes and disease. It has been 13 years since scientists first published the rough draft of the human genome and yet they are still just beginning to work out how our DNA influences our health. Most insights in recent years have come from so-called genome-wide association studies.

Full article is here:

http://www.nytimes.com/2013/11/28/science/linking-genes-to-diseases-by-sifting-through-electronic-medical-records.html?_r=1&

Here is the technical reference:

http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2749.html

It is easy to imagine a time when every child has their genetic sequence done at birth and that from there on a wide range of preventative and treatment possibilities will be able, over time, be able to become much safer and more evidence based.

New Interface Approaches (AI, NLP, Voice Recognition etc.)

It is widely recognised that data capture of information into electronic health records has been problematic ever since they were first developed decades ago. There are two key issues that need to be addressed. The first is the speed and ease of data capture and the second is the accuracy of the data that is captured. Additionally another factor which is always to be consideration on the impact of any computer use on the clinician workflow and patient comfort with the clinician / computer interaction.

That there is a problem can be seen from the following abstract:

Int J Med Inform. 2013 May;82(5):387-97. doi: 10.1016/j.ijmedinf.2012.08.004. Epub 2012 Sep 7.

A study of user requests regarding the fully electronic health record system at Seoul National University Bundang Hospital: challenges for future electronic health record systems.

Yoo S, Kim S, Lee S, Lee KH, Baek RM, Hwang H.

Source

Center for Medical Informatics, Seoul National University Bundang Hospital, Republic of Korea.

Abstract

OBJECTIVE:

Although the adoption rates for Electronic Health Records (EHRs) are growing, significant opportunities for further advances in EHR system design remain. The goal of this study was to identify issues that should be considered in the design process for the successful development of future systems by analyzing end users' service requests gathered during a recent three-year period after a comprehensive EHR system was implemented at Seoul National University's Bundang Hospital in South Korea.

METHODS:

Data on 11,400 service requests from end users of the EHR system made from 2008 through 2010 were used in this study. The requests were categorized as program modification/development, data request, insurance-fee identification/generation, patient-record merging, or other. The authors further subcategorized the requests for program modification/development into the following nine areas of concern: (1) indicators and statistics, (2) patient safety and quality of care, (3) special task-oriented functionalities, (4) ease of use and user interface, (5) system speed, (6) interoperability and integration, (7) privacy and security, (8) customer service, and (9) miscellaneous. The system users were divided into four groups--direct care, care support, administrative/insurance, and general management--to identify each group's needs and concerns.

RESULTS:

The service requests for program modification/development, data request, insurance-fee identification/generation, patient-record merging, and other issues constituted approximately 49.2%, 33.9%, 11.4%, 4.0%, and 1.5% of the total data set, respectively. The number of data-request service requests grew over the three years studied. Different groups of users were found to have different concerns according to their activities and tasks. Within the program-modification/development category, end users were most frequently concerned with ease of use and user interface (38.1% of the total) and special task-oriented functionalities (29.3% of the total) in their use of the EHR system, with increasing numbers of requests in both categories over the three years. Users in the direct-care group differed from the other groups in that they most frequently submitted requests related to ease of use and user interface, followed by special functionalities, patient safety and quality care, and customer service, while users in other groups submitted requests concerning ease of use and user interface and special functionalities with a similarly high frequency.

CONCLUSIONS:

Users have continued to make suggestions about their needs and requirements, and the EHR system has evolved to optimize ease of use and special functionalities for particular groups of users and particular subspecialties. Based on our experiences and the lessons we have learned in the course of maintaining full-EHR systems, we suggest that the key goals to be considered for future EHR systems include innovative new user-interface technologies; special extended functions for each user group's specific task-oriented requirements; powerful, easy-to-use functions to support research; new flexible system architecture; and patient-directed functions.

The abstract is found here:

http://www.ncbi.nlm.nih.gov/pubmed/22959193

There is a useful review of what vendor is the US are currently doing with EHR usability found here:

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.178.6190&rep=rep1&type=pdf

There is also a useful of some modern set of interfaces found here:

The 10 Best Designs in Medical Software

July 19, 2013

http://profitable-practice.softwareadvice.com/10-best-designs-in-medical-software-0613/

It is obvious that speed of use, ease of use and accuracy are vital. There are a range of technologies and approaches that may help.

First interface consistency - and ideally commonality across systems - can help in training and then ease of use. There is nothing more annoying than user interfaces that behave differently on different screens of the same system. (Even after a decade Windows still has a range of inconsistent foibles sadly).

Secondly in some situations pre canned reports that are able to be edited can speed up reporting - especially where most reports cover normal findings. (The risk is, of course, that such automation may mean reports are not properly considered and reviewed.)

Thirdly in some reporting situations - especially in radiology and pathology - voice recognition technology is already making a difference and as this is refined it can be expected the quality and efficiency will improve.

Beyond these technologies there are a range of more innovative ideas being considered and evaluated to try and make data capture faster, easier and more accurate. These include improved voice recognition, some types of artificial intelligence support and the use of natural language processing to improve the accuracy of diagnostic coding.

While discussing user interfaces it is important to bear in mind that the design of user interfaces must consider the risk of abuse by the use of such tricks and ‘cut and paste’ to fill in fields and save typing. Many studies have now shown this practice - and similar ones - are potentially dangerous. Clinicians need to be educated to not to use such practices.

Device / EHR Integration

A trend that has only recently emerged is the use of mobile devices to capture information and then link that information into a patient’s record via such technology as Wi-Fi or Bluetooth.

It is expected that there will be quite rapid growth in this sector over the next few years. This prediction appeared late in 2013. The article also explains the various benefits such integration can offer.

Big growth seen for device integration

Posted on Dec 02, 2013

By Mike Miliard, Managing Editor

The global medical device connectivity market, worth $3.5 billion this past year, is projected to top $33 billion by 2019, according to a new study.

The report, from Transparency Market Research expects a compound annual growth rate of 37.8 percent from 2013 to 2019.

Integration of data from medical devices into electronic medical records helps save time, eliminate transcription errors and improve overall care. But even as meaningful use EMR incentives augur big growth in device integration, the report points out that connectivity and operational issues, cost barriers for small and mid-sized providers and security concerns are all inhibiting the growth of this market.

Nonetheless, the market is on the upswing in a big way. It has been strongest in North America in recent years, thanks to increasing adoption of EMRs, and its continued growth is driven by factors such as increased need for workflow automation, efforts toward better patient safety, saved nursing hours, increased productivity of healthcare institutions, and minimizing the need for re-admissions.

The full article is found here:

http://www.healthcareitnews.com/news/big-growth-seen-device-integration

There is also some evidence that the use of such device / EHR integration (both in hospitals and with individual patients) can provide significant savings.

Medical Device, EHR Integration Could Save $30B: Study

Ken Terry

3/22/2013 10:17 AM

Interfaces must be standardized before hospitals can achieve financial, patient safety benefits, says West Health Institute report.

A new report from the West Health Institute (WHI) estimated that improving interoperability between medical devices and electronic health records (EHRs) in hospitals could save more than $30 billion a year while improving patient care and safety.

Among the sources of these savings, the report said, are increased capacity for treatment as a result of shorter lengths of stay ($18 billion); increased clinician productivity because of less time spent entering device data manually into EHRs ($12 billion); avoidance of redundant testing ($3 billion); and the reduction of adverse events because of "safety interlocks" ($2 billion).

Joseph M. Smith, MD, chief medical and science officer of WHI, presented the study in testimony before the House Energy and Commerce subcommittee. He also suggested regulatory and policy changes that he said could help create the conditions for interoperable systems that encompass medical device data. But in an interview with InformationWeek Healthcare, he said he hoped that the impetus for interoperability would come from the private sector.

The WHI report cited a recent study by HIMSS Analytics that said more than 90% of hospitals use six or more types of devices that could be integrated with EHRs. Examples include defibrillators, electrocardiographs, vital signs monitors, ventilators and infusion pumps. Yet only a third of hospitals integrate any medical devices with EHRs, and those that do, on average, integrate only three types of devices.

The full article is here:

http://www.informationweek.com/interoperability/medical-device-ehr-integration-could-save-$30b-study/d/d-id/1109208?

Managers can expect to see increasing requests for the procurement of new devices to be integrated with the various data capture and recording systems in use within the hospital. On the personal front simple capture of patient information to their PHR or to be able to download to their clinicians EHR can be expected to transform how many chronic illnesses are cared for and the quality of the information that is available to plan and review care.

Wearable Computing

2013 has been dubbed the year of ‘wearable computing’ with the increasing awareness of the possibilities of such devices really picking up in 2013.

A nice quick summary of the present scope is found in this article.

http://www.singularityhacker.com/post/44662876695/2013-the-year-of-wearable-computing

At present the most widely discussed applications are smart watches, activity monitors and, of course, Google Glass.

The short You Tube video which is linked here explains just what Glass is and what its present, but by no means final, capabilities are. I video also shows just how quickly Glass has evolved in 2 years from a truly messy prototype to a pretty refined product.

Play video from here:

http://youtu.be/V6Tsrg_EQMw

This is not to say that the idea of wearable computing and various extensions of the idea are at all new. It is really that there have been major practical difficulties which have only been surmounted recently.

A fictional example of an extreme example is seen in the screen shot of Captain Picard of The Enterprise as a Borg from Star Trek.

Source:

http://en.wikipedia.org/wiki/File:Picard_as_Locutus.jpg

There is a good recent article on the possible clinical applications of Glass here:

http://www.healthcare-informatics.com/blogs/gabriel-perna/2020-vision-google-glass-healthcare-coming

There is also a useful recent discussion from a Health Informatics Course on where all this might be heading longer term.

http://healthinformatics.wikispaces.com/Google+Glass

The use is training, supervision and documentation seem like obvious areas for use. As availability and functionality increases and cost comes down it will be fascinating to see where this eventually leads.

Similarly applications of smart watches and activity monitors (e.g. in fitness management etc.) might become interesting.

Lastly it is important to note that various implanted devices (hearing and optical implants) and machine / human interfaces are still in their infancy and can be expected to also evolve rapidly.

Artificial Intelligence In Decision Support

As talked about earlier in the course, effective interactive working Clinical Decision Support (CDS), is a major component expectation for potential benefits that arise from the use of electronic patient records. The vast majority of CDS that is presently implemented operate using embedded rules that interact with an information database. A typical example is where, as a prescription is written the medication prescribed is checked against the patient’s current therapy to make sure there are no potentially problematic interactions that might harm the patient. If potential problem is identified than the clinician is notified as the prescription is finalised.

Slightly more advanced system may similarly take advantage of already available physiological and diagnostic information to assesses any risks. An example might be to recommend a reduced dose of medications which are excreted by the kidney in those who have lesser renal function.

A key issue with all such rule based systems is that as rule complexity increases there are more risks of problematic or incorrect alerts. Additionally as more complex expectations are placed on such systems it becomes more and more difficult to create effective systems in a purely rules based way. Finally the subtlety and nuances of clinical information and treatment guidelines and publications has reached a level of complexity that simple rule based approaches simply do not work.

To now improve the capability and scope of CDS there needs to be delineation of improved and smarter ways of supporting decision making. To achieve this what is needed is the capability to handle both more information and less structured information contained in a range of different formats etc.

The range and scope of present CDS systems is quite usefully summarised in this summary from a HI Course in the US.

http://healthinformatics.wikispaces.com/Clinical+Decision+Support+Systems

The screen shots of the various systems and what they offer provides useful insight into the present state of play.

At the frontier at present was have the work being undertaken by IBM and a range of clinical partners.

This blog post provides useful links on where the ‘bleeding edge’ presently resides.

Understand IBM Watson – Bring Artificial Intelligence to Clinical Decision Support

February 26, 2013

IBM Watson is going to medical school at Cleveland Clinic. What Watson has to bring to medicine is the potential for advanced clinical decision support. Specifically algorithm-based, Bayesian decision analysis, rule based and expert systems. Several hurdles exist to accomplishing this: acquiring and validating of patient data, modeling of medical knowledge, keeping the data up-to-date, validate and integrate with the workflow. This process fits well with the Learning Healthcare System concept from the Institute of Medicine of taking research on evidence-based medicine into clinical decision support.

IBM Watson’s process in medical school will be to improve the inference graphs based on current data through human intervention. Providing clinical decision support is based on EMR data and the medical literature using DeepQA.

“The DeepQA project at IBM shapes a grand challenge in Computer Science that aims to illustrate how the wide and growing accessibility of natural language content and the integration and advancement of Natural Language Processing, Information Retrieval, Machine Learning, Knowledge Representation and Reasoning, and massively parallel computation can drive open-domain automatic Question Answering technology to a point where it clearly and consistently rivals the best human performance.”

Welcome Artificial intelligence to medicine and specifically clinical decision support.

The blog is found here:

http://ehealth.johnwsharp.com/2013/02/26/understand-ibm-watson-bring-artificial-intelligence-to-clinical-decision-support/

There is clearly a lot more to come in this area.

Robotics

It is well recognised that the operational costs of delivering healthcare are very much staff driven. Even with expensive high-tech equipment all over the costs of staffing still run to 70% or even more of a hospital’s budget. Overall Australia has 270,000 hospital staff (Source: AIHW - Hospital Statistics 2011-2012 ) and even with an average salary of $50,000 we are talking $13.5 billion p.a. of a total health spend of about $140 billion. Clearly there is an opportunity for cost reduction in such a large amount.

Indeed robotics would appear to be arriving in all sorts of situations.

The following links, to two slide shows, reveals the huge diversity of robotic applications that are now on offer.

http://www.informationweek.com/mobile/10-medical-robots-that-could-change-healthcare/d/d-id/1107696?

10 Medical Robots That Could Change Healthcare

12/6/2012 01:07 PM

Michelle McNickle

From microbots that scrape plaque from arteries to personal assistant robots that help care for patients, medical robots are transforming the face of healthcare.

http://www.informationweek.com/healthcare/clinical-information-systems/healthcare-robotics-patently-incredible-inventions/d/d-id/1111520?

Healthcare Robotics: Patently Incredible Inventions

9/12/2013 02:33 PM

Onat Ekinci

Medical robots will change the operating room much like PCs reshaped the office. Get an advance look from these cutting-edge robotic technology patents and patent applications.

What presently look to be the most important emerging applications are robotically assisted surgery, dispensing robotics (to reduce dispensing errors), the emerging area of automated delivery of supplies to wards (and sample delivery to labs) and various assistive robots to help care for the aged and infirm.

This is certainly another area with a good way to go especially as there are improvements in artificial intelligence systems to improve capabilities.

Mobile Health and Associated Technologies

Section 7 of the course has a discussion on both the present and potential future state of the interaction between mobility, devices and healthcare delivery. There is not a day that goes by that new apps in the health domain and not announced, that great ideas are coming to the market and that we hear from regulators regarding their concerns about the quality and value of what is on offer.

As with most of what is seen above this area is also a moving field - but it needs to be mentioned here to remind readers just how dynamic the area of mHealth (for mobile health) is and how it is important for managers to remain alert to what is going on.

3D Printing

One of the most exciting innovations of the last few years has been the quite rapid development of the technology of 3D printing.

Wikipedia has a useful description:

Additive manufacturing or 3D printing is a process of making a three-dimensional solid object of virtually any shape from a digital model. 3D printing is achieved using an additive process, where successive layers of material are laid down in different shapes. 3D printing is also considered distinct from traditional machining techniques, which mostly rely on the removal of material by methods such as cutting or drilling (subtractive processes).

A materials printer usually performs 3D printing using digital technology.”

The full article is here:

http://en.wikipedia.org/wiki/3D_printing

Of interest in this section is the use of 3D printing techniques in healthcare - with applications in everything from prosthesis manufacturing to tissue engineering.

Where this is all heading is covered in a fascinating article in the UK Financial Times. Here is the link:

http://www.ft.com/intl/cms/s/0/74c5d5b6-4b9a-11e3-8203-00144feabdc0.html?ftcamp=crm/email/20131115/nbe/DrugsHealthcare/product&siteedition=intl

Again it is by no means where this technology will go over the next few years.

Health Systems Science

A theme of this course has been to emphasise the need where possible to apply evidence to assist in making appropriate clinical, managerial and policy decisions.

A limitation in following this approach is that many policy issues and problems can be described as ‘Wicked Problems”

Wikipedia defines such problems thus:

"Wicked problem" is a phrase originally used in social planning to describe a problem that is difficult or impossible to solve because of incomplete, contradictory, and changing requirements that are often difficult to recognize. The term "wicked" is used to denote resistance to resolution, rather than evil. Moreover, because of complex interdependencies, the effort to solve one aspect of a wicked problem may reveal or create other problems.”

See here:

http://en.wikipedia.org/wiki/Wicked_problem

Most in the health sector will recognise much of what they do as having to be based in information which is “incomplete, contradictory, and where changing requirements that are often difficult to recognize”.

Additionally there are often elements of complexity and interaction between various actors that cannot simply be resolved.

One approach to addressing such apparently insoluble issues and problems is through the use of dynamic modelling techniques that can capture the complexity and interaction of the various issues at play and provide a way of examining them, which while not being perfect can be useful and permit improved understanding and testing of policy options.

Dynamic modelling techniques are not new - but in the last decade improved modelling software and more powerful computers has provided the opportunity to apply these very powerful techniques more easily.

A recent thesis from South Africa looking at some issues surrounding approaches to HIV gives some insights regarding a what is possible:

http://upetd.up.ac.za/thesis/available/etd-06062013-122711/unrestricted/dissertation.pdf

At present such techniques are almost certainly underutilised in evolving a more evidence based, effective and safe health system.

Another Perspective On A Similar Theme

I found this listing in my research and think it is a useful set of alternative perspectives on essentially the same topic to consider.

http://www.healthcare-informatics.com/article/top-ten-tech-trends-2012-time-exhilaration-and-anxiety

Top Ten Tech Trends 2012: A Time of Exhilaration and Anxiety

February 28, 2012 by The Editors

2012 Top Ten Tech Trends

Performance Imperatives

Trend: Performance Measurement

By Mark Hagland

Population Health Management and Readmissions

Trend: Population Health Management

By Gabriel Perna

Turning Healthcare's Business Model Inside Out

Trend: ACOs and Care Coordination Tools

By Jennifer Prestigiacomo

Bridging the Care Transition Gap

Trend: Care Management Transitions

By John DeGaspari Second-Generation Clinical Decision Support

Trend: Decision Support

By Mark Hagland

Year of the CISO

Trend: Privacy and Security

By David Raths

Private HIEs on the Upswing

Trend: Private vs. Public HIEs

By Jennifer Prestigiacomo

Imaging Informatics and the Enterprise

Trend: Imaging

By Gabriel Perna

The BYOD Revolution

Trend: Mobile Health

By David Raths The Game Changer

Trend: Personalized Medicine

By David Raths

—The Editors

There is nothing to be lost by reviewing these for another view

Last Words On The Overall Course

Overall this short unit has had two objectives. The first is to introduce readers to what is a really interesting and potentially invaluable part of health care delivery going forward and one that has not really been successfully fully exploited to date. The second has to been to warn that there is no such thing as a ‘magic pudding’ and that the benefits that can be gained from Health IT are neither easy, inevitable or painless. If you come away with that perspective I will be greatly satisfied.

Review Questions

1. Considering the various futures considered in this Section which three do you expect to have the largest positive impact over the next decade?

2. Considering the Health Sector as a whole, which aspect presently provides the most difficult ‘wicked problem’ for managers and policy managers to address?

Return to HI Homepage

Questions & Comments to Geoff McDonnell
Personal tools