InPredictive Modeling News’slatest issue, their editors interview our CEO about analytics and end-of-life care and highlight our newest client, Doctor’s Choice. Check it out below.
Analytics Paves Way for Compassionate Conversations
Understanding the probability of death informs patient choices
Analytics can help providers objectively understand which patients would benefit most from an end-of-life conversation — which will now be reimbursed by Medicare. Predictive analytics technology accomplishesthat by identifying patients likely to die in the next 90 days, so clinicians can delve deeper into their records to further assess the situation.
After reviewing the analytics, the clinician and patient’s physician can have a conversation with the patient about his or her end-of-life care options, such as hospice care. Predictive Modeling News talked to Dan Hogan, founder of and CEO at Medalogix, a Nashville-based predictive analytics provider for home and hospice care agencies.
Predictive Modeling News:When does Medicare reimbursement kick in? Is this something providers need to jump on right now?
Dan Hogan:CMS finalized the advanced care planning reimbursement rule as part of its Medicare physician fee schedule for calendar year 2016. The rule will take effect January 1; it reimburses physicians for end-of-life care planning conversations. While physicians don’t need analytics tools to determine whom to have end-of-life planning conversations with, our end-of-life analytics tool, Medalogix Bridge, can help pinpoint which patients could most benefit from those conversations — those patients who are likely to pass away within 90 days.
PMN:What kinds of markers does the analytics program look for when determining who’s likely to die in the next 90 days?Specific diagnoses? Specific trends in clinical lab results?
DH:Interestingly enough, diagnosis is not regularly predictive. However, severity of diagnosis is regularly predictive. Our Number Two patient mortality risk factor is actually ulcers. Ulcers are statistically predictive of patient mortality. Another predictor — or non-predictor, really, which most care providers are surprised by — is that patients who move to home health from the hospital are not sicker and therefore not more likely to pass away. Most are under the impression that patients from the hospital are more likely to pass away — and that’s not true.
PMN:What other factors go into the analytics besides likelihood of death? Do religious or family preferences figure in? Or is that part of the discussion that follows?
DH:The product of the model is a probability of death. As you noted though, the discussion is the most important part. This is where patients’ choices come in, which includes religious and family preferences. And that’s how it should be. Our technology helps navigate not who should go to hospice — only the patient can decide if hospice is right for him or her — but who would benefit from the end-of-life conversation. Our technology is just the starting point. The clinicians, the patients and their families make the hospice decision.
PMN:Walk us through a sample use of the analytics program. How often does it scour records for information about likelihood of death? What kind of report does a user of the analytics program receive?
DH:Bridge, our analytics and workflow technology solution, leverages predictive analytics from EMR data and identifies which patients on the home health census could benefit most from hospice care. Bridge ranks patients according to their relative appropriateness for hospice, which allows clinicians to easily view which patients have the highest risk ranking. Then, Bridge provides a built-in workflow to help clinicians organize the necessary steps from patient identification to having a conversation with the patient about his or her end-of-life care options.
PMN:What kinds of patterns the human eye can miss does the analytics program discover? What would a user of the program look for?
DH:These are patterns that just by sheer quantity are not apparent to the human eye. There is no way even a 30-year veteran clinician can draw on the experience of millions of patients over decades. Computers and predictive analytics can. This information is not meant to replace the human eye or the clinician’s experience, but rather to augment it. I like to think of it as four dimensions of decision making. Traditionally we use our experience, instinct and education to make decisions. That’s what clinicians use when they’re making patient care decisions. Analytics adds a fourth dimension to this decision making process. With another dimension, you can see a clearer picture and generate better results.
PMN:What happens if the data don’t crunch properly and a name comes up that shouldn’t? Is there a fail-safe mechanism to keep providers from having a difficult conversation with the wrong patient?
DH:Yes. Clinicians are the fail-safe mechanism. Our technology is not the end. It is not the decision maker. It is created to deliver statistically significant insights to clinicians so they have more information to direct important care decisions. Our technology is designed that way because there are some things, like human will, that can’t possibly be accounted for in a predictive model. For instance, if Mrs. Jones’ granddaughter is going to get married in a year, her clinician knows that Mrs. Jones has something she wants to live for. It’s at that point Mrs. Jones’ clinician looks at our stratification and says, “Heck no, Mrs. Jones is not going to die.” Because this clinician has first-hand personal information, she won’t put Mrs. Jones in the workflow to have the end-of-life conversation.
PMN:What other ways can analytics be used in home health and hospice care?
“Interestingly enough, diagnosis is not regularly predictive. However, severity of diagnosis is regularly predictive.”
DH:Analytics can also be used to automate a home health clinical team’s touch points. For example, our solutionTouchenables clinicians to quickly see which patients are most at risk for transferring off census before their 60-day care episode completes. Then scheduling touch point calls can begin, whether manually or through Interactive Voice Response. Not only will this save hours from a clinician’s day, but it will also improve patient care and an agency’s bottom line. Analytics can also add insights and efficiency to home health providers’ end-of-episode planning and post-discharge patient calling programs. Our solutionNurturecentralizes, streamlines and expedites the programs and also catalogs calls and referrals so all information is easily managed.
PMN:What other analytics applications in home and hospice care will we see in the future?
DH:We’re working on a predictive modeling tool for home health value-based purchasing. It’ll be more granular in terms ofrisk. We’re building models that can predict the probability of a patient thriving in each process measure or category.
Doctors’ Choice Home Health Newest Medalogix Client
Medalogix, a home health and hospice analytics company, reports its newest partner in care,Doctors’ Choice Home Health, a certified Medicare- and ACHC-accredited home health agency serving Northeast Florida. Doctors’ Choice is deploying Medalogix’s analytics-based population health management solution, Touch, a statement points out, which “helps home health providers increase touch points with the most at-risk patients to avoid unnecessary readmissions and drive referrals.” In a peer-reviewed study, the statement adds, the technology solution “proves to reduce readmission by a relative rate of more than 20%.”
Says Kathy Edwards RN, Director of Operations at Doctors’ Choice Home Health: “Medalogix is a wonderful analytics and workflow solution that allows us to focus on increasing touch points with patients who are at risk for readmission.” Adds Dan Hogan, CEO at Medalogix: “We’re proud to work with Doctors’ Choice in its charge to leverage technology to reduce hospital readmissions and improve patient care. They’re a progressive team dedicated to providing the best in home healthcare.”
Medalogix offers two home health solutions: Medalogix Touch and Bridge. Touch is a population health management technology that identifies patients who are at the greatest risk of transferring off home health census and those who would benefit from an additional care episode. After identifying the relevant patients, Touch equips providers with an automated calling component that schedules appropriate follow-up calls and a watch list module that helps monitor high-risk patients nearing the end of their episodes. Bridge identifies patients who could benefit from hospice care and then equips providers with organizational workflows to ensure the right patients hear about their end-of-life options at the right time. The solution is especially useful for dual home health and hospice providers, the statement says, adding predictive insights and operational efficiency to their hospice transition process. The Bridge solution was recently recognized by Harvard University as an impactful healthcare innovation when Medalogix was selected from a pool of 500 applicants as a finalist in its Health Acceleration Challenge.
See the full publicationhere.