EditorialPublication Exclusive

Small molecules, ‘big data’ and the EMR

It’s strange how apparently unrelated events sometimes seem to converge into a common theme.

During the past couple weeks, I attended a preview of — and subsequently watched in its entirety — the remarkable Ken Burns documentary Cancer: The Emperor of All Maladies.

I spent a couple of days at a meeting focused on cancer genomics and proteomics and the potential power of “big data.” Upon returning home, I have been in meetings to plan the next upgrades to our electronic medical record (EMR) and have had some discussions about potential add-ons to our EMR to meet some of our needs in terms of clinical care pathways, quality metrics and so on.

The ‘really big’ picture

John Sweetenham, MD, PhD

John Sweetenham

How do these events come together in my mind?

Let’s start with the EMR. When I hold medical staff or program leaders’ meetings at our institution, one subject always guaranteed to provoke animated discussion is our EMR. The challenges and frustrations of using the EMR, the intrusion on work flow and the additional work generated — especially considering we transitioned to a new EMR within the last year — are major sources of discontent among providers.

My colleagues fully understand the necessity of the EMR in terms of patient and practice management in the 21st century, but the day-to-day clinic issues such as documentation, chemotherapy orders, scheduling and billing using the EMR are very real and can make it difficult for us to embrace the bigger picture.

What is becoming increasingly apparent is that the bigger picture is really big! The EMR likely will become much more than a record. It is going to become one of the most important — if not the most important — tools we use in oncology practice, not just because of the intrinsic functionality, but because it will become the window into the world of big data, pathway-driven treatment and real-time, evidence-based clinical decision support.

The number of big data initiatives in the oncology world is increasing rapidly. ASCO has given a great deal of exposure to its CancerLinQ project in recent weeks. The Oncology Research Information Exchange Network (ORIEN) recently expanded its membership from two academic medical centers to six. Like many similar initiatives, this project seeks to link clinical and demographic data to tissue and DNA with the aim of furthering the progress toward precision oncology. Many other private companies, academic medical centers and collaborative groups are developing similar programs.

All of these efforts are chasing the ultimate goal of building comprehensive genomic/proteomic databases, correlated with clinical outcome data that will allow us to identify specific pathways for targeted therapies in individual patients, or even identify the most effective chemotherapy regimens when small-molecule targeted therapies are not relevant.

As I sit at my desktop in the clinic, trying to find a way to modify a chemotherapy dose or change an order within the medical record, the lofty goal of precision oncology seems a long way off. We could be forgiven for thinking that real-time, biologically based decision support is so far into the future that we can only dream about it as a remote possibility. After the events of the last couple of weeks, however, I am now starting to believe that this will be a clinical reality for us much sooner than most of us might have expected.

A new reality

So how might this look?

One vision of the ideal situation would be that having taken a biopsy and established a cancer diagnosis, that biopsy goes through genomic/proteomic analysis (personally, I’m out of my comfort zone on the technical details here, but as a practicing oncologist, I don’t think I’m alone in this respect).

When I see the patient in the clinic, through the EMR I will have access to a report that will — in addition to the conventional diagnostic and staging data — provide me with information about clinically relevant genomic or proteomic alterations. It will also let me know whether there are available drugs or clinical trials for drugs that target that pathway, or whether there are chemotherapy options that should be used. These data will be derived from the aggregated big data sets.

Using this information, I can — still at my desktop — explore trial options or select a therapy from those identified in the report. Having selected a therapy, I will be able to determine the clinical evidence base for that approach and, at the same time, examine my center’s agreed care pathways to determine whether my preferred choice is consistent with our chosen pathway(s).

The system also will link to the patient’s insurance company to request pre-authorization electronically. The decision support on the payer end can be made using the same evidence-based tools that I used to select therapy. The treatment then can be approved in real time, and I can move to a screen that will populate with the treatment order.

Every time we go through this process, the entire system accumulates more data, aggregates the outcome data and learns. So, through the window of the electronic record, I have:

  • gained access to a real-time, continually learning database;
  • chosen an evidence-based therapy or clinical trial;
  • gained (hopefully) insurance approval for that approach; and
  • have written the order.
  • Subsequent follow-up will keep adding to this knowledge base.

    Even a couple of years ago, this would have sounded unlikely. Now, I think this is likely to be a reality within the next few years.

    A worthwhile effort

    So where does Cancer: The Emperor of All Maladies fit into all of this?

    The documentary takes us through the history of our understanding of cancer biology and treatment, highlighting many of the landmark discoveries and trials that have changed our thinking, revolutionized treatments and improved the lives of many affected by cancer.

    The big data/precision oncology effort has the potential to transform our thinking and fundamentally change cancer care delivery. Anatomic/organ origin of a tumor may become much less relevant to treatment — at least after surgery — than altered genes or proteins, and treatment selection will be based on truly real-time data.

    The EMR — and the access it provides to massive data sets and decision support — will be the most powerful tool a practicing oncologist has ever had.

    Of course, this vision is based on certain assumptions, not the least of which is that we will have the bioinformatics capability to correctly interpret the masses of genomic and proteomic data.

    Also, we may need to change our thinking on what constitutes level one evidence. New clinical trial methodologies plus the power of individual patient data will need to be validated in the real world and may (or may not) prove equivalent.

    The next time you are struggling with e-prescribing, inserting text, writing an inpatient order or trying to find outside records that have been scanned into the EMR, think of how different this all could be within the next few years. It may seem unlikely now, but I think it’s going to be worth all the frustration and effort — really.

    For more information:

    John Sweetenham, MD, is HemOnc Today’s Chief Medical Editor for Hematology. He also is senior director of clinical affairs and executive medical director at Huntsman Cancer Institute at the University of Utah. He can be reached at john.sweetenham@hci.utah.edu.

    Disclosure: Sweetenham reports no relevant financial disclosures.

    It’s strange how apparently unrelated events sometimes seem to converge into a common theme.

    During the past couple weeks, I attended a preview of — and subsequently watched in its entirety — the remarkable Ken Burns documentary Cancer: The Emperor of All Maladies.

    I spent a couple of days at a meeting focused on cancer genomics and proteomics and the potential power of “big data.” Upon returning home, I have been in meetings to plan the next upgrades to our electronic medical record (EMR) and have had some discussions about potential add-ons to our EMR to meet some of our needs in terms of clinical care pathways, quality metrics and so on.

    The ‘really big’ picture

    John Sweetenham, MD, PhD

    John Sweetenham

    How do these events come together in my mind?

    Let’s start with the EMR. When I hold medical staff or program leaders’ meetings at our institution, one subject always guaranteed to provoke animated discussion is our EMR. The challenges and frustrations of using the EMR, the intrusion on work flow and the additional work generated — especially considering we transitioned to a new EMR within the last year — are major sources of discontent among providers.

    My colleagues fully understand the necessity of the EMR in terms of patient and practice management in the 21st century, but the day-to-day clinic issues such as documentation, chemotherapy orders, scheduling and billing using the EMR are very real and can make it difficult for us to embrace the bigger picture.

    What is becoming increasingly apparent is that the bigger picture is really big! The EMR likely will become much more than a record. It is going to become one of the most important — if not the most important — tools we use in oncology practice, not just because of the intrinsic functionality, but because it will become the window into the world of big data, pathway-driven treatment and real-time, evidence-based clinical decision support.

    The number of big data initiatives in the oncology world is increasing rapidly. ASCO has given a great deal of exposure to its CancerLinQ project in recent weeks. The Oncology Research Information Exchange Network (ORIEN) recently expanded its membership from two academic medical centers to six. Like many similar initiatives, this project seeks to link clinical and demographic data to tissue and DNA with the aim of furthering the progress toward precision oncology. Many other private companies, academic medical centers and collaborative groups are developing similar programs.

    All of these efforts are chasing the ultimate goal of building comprehensive genomic/proteomic databases, correlated with clinical outcome data that will allow us to identify specific pathways for targeted therapies in individual patients, or even identify the most effective chemotherapy regimens when small-molecule targeted therapies are not relevant.

    As I sit at my desktop in the clinic, trying to find a way to modify a chemotherapy dose or change an order within the medical record, the lofty goal of precision oncology seems a long way off. We could be forgiven for thinking that real-time, biologically based decision support is so far into the future that we can only dream about it as a remote possibility. After the events of the last couple of weeks, however, I am now starting to believe that this will be a clinical reality for us much sooner than most of us might have expected.

    A new reality

    So how might this look?

    One vision of the ideal situation would be that having taken a biopsy and established a cancer diagnosis, that biopsy goes through genomic/proteomic analysis (personally, I’m out of my comfort zone on the technical details here, but as a practicing oncologist, I don’t think I’m alone in this respect).

    PAGE BREAK

    When I see the patient in the clinic, through the EMR I will have access to a report that will — in addition to the conventional diagnostic and staging data — provide me with information about clinically relevant genomic or proteomic alterations. It will also let me know whether there are available drugs or clinical trials for drugs that target that pathway, or whether there are chemotherapy options that should be used. These data will be derived from the aggregated big data sets.

    Using this information, I can — still at my desktop — explore trial options or select a therapy from those identified in the report. Having selected a therapy, I will be able to determine the clinical evidence base for that approach and, at the same time, examine my center’s agreed care pathways to determine whether my preferred choice is consistent with our chosen pathway(s).

    The system also will link to the patient’s insurance company to request pre-authorization electronically. The decision support on the payer end can be made using the same evidence-based tools that I used to select therapy. The treatment then can be approved in real time, and I can move to a screen that will populate with the treatment order.

    Every time we go through this process, the entire system accumulates more data, aggregates the outcome data and learns. So, through the window of the electronic record, I have:

  • gained access to a real-time, continually learning database;
  • chosen an evidence-based therapy or clinical trial;
  • gained (hopefully) insurance approval for that approach; and
  • have written the order.
  • Subsequent follow-up will keep adding to this knowledge base.

    Even a couple of years ago, this would have sounded unlikely. Now, I think this is likely to be a reality within the next few years.

    A worthwhile effort

    So where does Cancer: The Emperor of All Maladies fit into all of this?

    The documentary takes us through the history of our understanding of cancer biology and treatment, highlighting many of the landmark discoveries and trials that have changed our thinking, revolutionized treatments and improved the lives of many affected by cancer.

    The big data/precision oncology effort has the potential to transform our thinking and fundamentally change cancer care delivery. Anatomic/organ origin of a tumor may become much less relevant to treatment — at least after surgery — than altered genes or proteins, and treatment selection will be based on truly real-time data.

    The EMR — and the access it provides to massive data sets and decision support — will be the most powerful tool a practicing oncologist has ever had.

    Of course, this vision is based on certain assumptions, not the least of which is that we will have the bioinformatics capability to correctly interpret the masses of genomic and proteomic data.

    Also, we may need to change our thinking on what constitutes level one evidence. New clinical trial methodologies plus the power of individual patient data will need to be validated in the real world and may (or may not) prove equivalent.

    The next time you are struggling with e-prescribing, inserting text, writing an inpatient order or trying to find outside records that have been scanned into the EMR, think of how different this all could be within the next few years. It may seem unlikely now, but I think it’s going to be worth all the frustration and effort — really.

    For more information:

    John Sweetenham, MD, is HemOnc Today’s Chief Medical Editor for Hematology. He also is senior director of clinical affairs and executive medical director at Huntsman Cancer Institute at the University of Utah. He can be reached at john.sweetenham@hci.utah.edu.

    Disclosure: Sweetenham reports no relevant financial disclosures.