Readers Write: Navigating the Early Days of Healthcare AI Integration

Navigating the Early Days of Healthcare AI Integration
By Michael Burke

Michael Burke, MBA is founder and CEO of Copient Health of Atlanta, GA.

Have you tried using any of the AI tools that have taken the world by storm recently? This article will probably be more helpful if you have some knowledge or experience with ChatGPT, Google Bard, Anthropic Claude, or any other LLM/chat model tool.

If you haven’t already, try asking one of these tools a specific question or give it an assignment to produce a specific document and see where it leads. You may be surprised at just how useful the results can be.

If you’ve used these tools to answer questions or generate content (e.g., a legal document, a policy document, an email, or an article like this one), you have some sense of their potential. Imagine what could be done with a tool that leveraged an LLM like ChatGPT on your hospital’s data. The software vendors you use are all either investigating or actively releasing tools powered by LLMs to leverage your data. At Copient Health, we are, too.

It’s my belief that these tools will fundamentally change the way you interact with those vendor systems and ultimately, in both the way that you do your work and the results that you get.

A comprehensive list of all use cases is impossible because we’re so early in the process, but here are a few obvious low-hanging fruit uses that are relevant for software vendors:

• LLMs are already powering chart notes that are built in real time from patient conversations.
• Dashboards and reports will become unnecessary, because you will always have the specific data or chart that you need just a query away. The LLM can even proactively push the appropriate information in the appropriate format for the appropriate context.
• You can forget about manuals, indexed help systems, or frustrating first-generation chat bots that perform poorly. LLM-powered solutions are better at finding what you’re looking for using a similarity search of a vector database.
• You might even abandon memorizing complex commands or menu hierarchies and ask the LLM to accomplish the task instead.

But ChatGPT and other public-facing LLMS were trained on public data. How can they be leveraged for use cases that require knowledge of private data?

The answer to that question used to take a lot of time, money, and a team of data scientists to train your own LLM, or at least fine tune an existing open source model. That has changed dramatically, mostly in the last 6-8 months, based primarily on a term that you may have heard: “prompt engineering,” and one that you probably haven’t: “in-context learning.” Here’s a quick summary:

LLM models are text-in, text-out black boxes. But the text-in doesn’t have to be limited to a simple question. It can include prompts of background information, examples of questions and answers to similar scenarios, chunks of data, or simply directing the LLM to “think step-by-step.”

These are all basic forms of prompt engineering. The LLM temporarily “learns” from this prompt data, at least enough for your current conversation. LLMs can be used as an inwardly-directed service to decide what data or tool to use based on the prompts that it receives. This design pattern has demonstrated better results than the more cumbersome fine-tuning approach for the smaller data sets that we’re talking about.

An entire ecosystem of software tools has emerged to support the use of these pre-trained LLMs on private data. These tools convert the challenge from what was once an arcane AI data science problem to a data engineering problem, primarily built around prompt engineering and in-context learning.

Here’s an illustration of how quickly these tools have evolved and been adopted. One of the most widely used tools in the ecosystem, LangChain, was first introduced in October last year as an open source project from two college students. In a few months, its use expanded globally. The founders incorporated and raised $20 million in venture funding from Sequoia Capital. Since last October, they have garnered 60,000 GitHub stars, which is a measure of its popularity among software developers. For context, Python, the language the LangChain toolset is written in, has fewer stars over a significantly longer time period: 51,500 stars over six years. ChatGPT itself captured one million users in just five days.

This head-spinning rate of change gives an advantage to startups, given their rapid iteration and integration of new tools and ideas. Some large healthcare software vendors that are infamous for relying almost exclusively on internally developed tools find themselves in a challenging situation. It’s impractical for them to build their own LLMs, as they would likely never rival the performance of commercially available options, and it would take forever. And since they are not used to relying on third-party software as part of their solution, they aren’t prepared for the rate of change at which these solutions are evolving.

For instance, just yesterday, LangChain had 18 separate commits (i.e., changes) to their codebase. That’s fast! Adapting to rapid changes and advancements requires a new level of agility.

We’ve recently heard announcements and partnerships from big tech and big healthcare IT. It will be interesting to see if these announcements produce real value in the near term, or if they are just a way to buy time for the vendor to figure out this rapidly evolving space.

Readers Write: It’s Time for a National Patient Identifier

It’s Time for a National Patient Identifier
By Gregg Church

Gregg Church is president of 4medica of Marina del Rey, CA.

Congress has the power to make healthcare safer and less expensive for patients, payers, and providers. It can do this by removing the ban that prohibits using federal funds for the development of a unique national patient identifier.

The patient identifier system would give each patient a single ID that would follow them through their healthcare journey, regardless of provider or payer, while still protecting their private information. It would reduce medical and billing errors and denied claims, while eliminating countless hours insurers and hospital systems spend resolving patient matching errors. It would also aid medical research and make it easier for our healthcare system to respond effectively during national emergencies, like the COVID-19 pandemic.

The federal ban on a national patient identifier was born from good intentions. Former US Rep. Ron Paul in 1998 added the provision to the Labor-HHS appropriations bill. The physician and libertarian cited concerns about patient privacy and the dangers of the federal government collecting and centralizing medical records. His son, Sen. Ron Paul, also a physician and libertarian, now leads the opposition with the help of the ACLU and other groups.

Congress has come close to ending the ban. For the past four fiscal years, the House has removed it from its version of the appropriations bill; the Senate did likewise the past two years. Each year, however, it has been reinstated in the final budget.

In 2021, Patient ID Now, a coalition of more than 40 healthcare organizations, including the American College of Surgeons, American Heart Association, American College of Cardiology, The Joint Commission, and American Health Information Management Association, was formed to push for a nationwide strategy to address patient identification.

The group noted that the ban was put in place 25 years ago at a time when patient records were still largely kept in manila folders. It’s now a hindrance to the necessary digitization of healthcare. While concerns over patient privacy are real, a national patient identifier could be implemented in such a way that it protects patients.

I like to believe that much of the opposition to a universal patient identifier is due to a lack of awareness of the volume of incomplete, duplicate, missing, and overlaid medical records and the problems they cause.

Imagine if your personal finance records had a roughly one in five chance of being duplicated or mixed up with someone else’s accounts by financial institutions. Think of the chaos and damage and the ensuing demands to fix the problem.

Duplication of patient records is one of the most serious problems with healthcare data quality, and it’s more common than many think. Duplication rates are as high as 30% in some healthcare organizations, and a 10% rate is common. Up to half of patient records are not matched in transfers between healthcare systems. 

Patients are endangered by low-quality records, particularly duplicate and overlaid records, in which the data for two patients is mixed.

Approximately 70% of care decisions are based on lab tests, which are performed by techs working in relative isolation from the care team. Labs frequently create duplicate records while entering patient information into computers. That bad data can then be multiplied and disseminated throughout a hospital system and between systems.

Clinicians working from bad data can misdiagnose, prescribe the wrong course of treatment, and order duplicate tests, which delay necessary treatment.

Black Book in 2018 surveyed health technology managers about problems with patient identification processes. It found that the cost of medical care due to duplicate records averaged $1,950 per patient per inpatient stay and more than $800 per ED visit.

Those surveyed also estimated that 33% of denied claims were due to inaccurate patient identification or information. That cost the average hospital $1.5 million in 2017 and the US healthcare system more than $6 billion annually.

That unnecessary expense could be eliminated with a standard patient identification system.

True interoperability among patients, providers, and payers is a goal of the healthcare industry, one that could be made more achievable through a patient identifier system that allows for the disruption-free exchange of patient records.

While recent improvements in patient identification processes, such as hospitals adopting Enterprise Master Patient Indexes and the use of machine learning, have improved record matching, it’s barely keeping pace with the explosion in medical records and the sharing of data among different healthcare organizations.

Adopting a national patient identifier would be a significant step toward building a safer, and more effective and affordable healthcare system. It’s time for Congress to listen to the experts and remove the ban. We’ll all be better off for it.

Readers Write: Tell Me Again Why Fax is Superior?

Tell Me Again Why Fax is Superior?
By Dan Wilson

Dan Wilson is founder and CEO of Moxe Health of Madison, WI.

The “inherent security” provider argument in a recent KLAS report on digital fax is unfounded and a remnant of another era. User error is inherently possible when a process involves manual steps, and fax isn’t secure relative to more modern ways to encrypt and transact data between multiple parties. It’s also unlikely that faxing remains analog (using only a phone line), as groups are using Efax or VOIP lines with transactions going over the Internet and the fax isn’t actually encrypted. 

“Ease of use” sounds like the person who used to say that “no one will text, because it’s easier to just pick up the phone and call someone.” Faxing is easy only because finding a directory of where to send files electronically is so hard. If we solve the directory issue, the “ease of faxing” benefit is reduced.

Another way to think about ease of use is that it’s actually a tradeoff for security. Fax is easy because you send a document to a clinic’s single number. That means that the message isn’t specific to a patient or recipient. Anyone who has access to the fax machine can see the information. Rarely do you get both ease of use and security, but there’s a better set of options with digital exchange to select the right tradeoffs based on the sensitivity of the information versus just having a blunt tool.

Fax is hopelessly outdated. It creates enormous manual effort and adds cost on both ends of the transaction. A CAQH study estimates that faxing or mailing instead of using digital transfer costs $25 billion per year.

For the love of God, can we stop making doctors do a ton of work to digitize records and paying people to print them and fax them, taking those records from digital to analog and then to an even worse version of analog (an image)? And then consuming massive resources on the recipient’s end to try to reconstitute a digital copy of what started its life as a digital record? And along the way, losing fidelity of information in addition to people and compute time.

Tell me again why fax is superior?

Readers Write: What’s Needed to Resolve the Medicaid Redetermination Crisis

What’s Needed to Resolve the Medicaid Redetermination Crisis
By Carrie Kozlowski

Carrie Kozlowski, OT, MBA is co-founder and COO of Upfront Healthcare of Chicago, IL.

More than 90 million Americans, including children, the elderly, people with disabilities, and veterans rely on Medicaid for their health coverage. Many of them are now losing that coverage with the expiration of the COVID-19 relief laws this spring.

According to the Kaiser Family Foundation (KFF), as of mid-July, more than two million people have lost Medicaid coverage since April 1, 2023, with most of them removed from state rolls for technicalities, such as missing the deadline to complete their forms or to file certain required documents. The Centers for Medicare and Medicaid Services’ (CMS) also found that 31% of renewals resulted in someone dropping Medicaid or Children’s Health Insurance Program (CHIP) benefits, with an alarming 79% of the beneficiaries losing coverage due to procedural reasons, not because they no longer qualified due to income or changes in family arrangements.

In total, KFF estimates that 15 million people will be dropped from Medicaid over the course of the year under this “Medicaid unwinding” process. The result? Healthcare enterprises will experience a gap in compensation for care, operational efficiency will suffer, and patients will get sicker during this post-pandemic Medicaid redetermination period.

The federal government is stepping in to try to stem the crisis. On July 19, CMS reported that it has intervened with several states, requiring them to pause procedural terminations and reinstate individuals. Moving forward, the CMS will be closely tracking state data and fielding complaints to identify problems early with renewals and take corrective action, according to the fact sheet “Returning to Regular Medicaid Renewals: Monitoring, Oversight, and Requiring States to Meet Federal Requirements” released by the agency.

More efforts are needed, however. Basic lack of awareness about the changes in the laws is a key part of why the Medicaid unwinding process is turning into a crisis in many states. A Robert Wood Johnson Foundation survey, “Awareness of the Resumption of Medicaid Renewal Processes Remained Low in December 2022,” revealed that approximately 64% of Medicaid members had heard nothing at all about the enrollment requirements, leaving them vulnerable to losing their coverage.

All this is a significant concern, not only from a population health and health equity perspective, but it also because it has far-reaching financial implications for health systems and medical group that are already facing slim to negative operating margins. With declining enrollees, they risk further negative financial impacts and may need to increase staffing to facilitate point-of-care enrollment, adding to the costs and inefficiencies.

It is crucial for health systems to keep patients enrolled in Medicaid, not just for the sake of their health, but for the financial stability of their own operations. Keeping them enrolled ensures that they can continue to receive preventative care, which leads to improved health outcomes, protected reimbursement, and reduced overall healthcare costs.

From an operational standpoint for health systems, it is also in their best interest to keep Medicaid patients covered so they do not lose access to primary care providers, causing delayed time to treatment and sicker patients admitted to hospitals, flooding intensive care units, and causing backlogs in emergency departments that can reverberate through the hospital and can delay elective surgery schedules.

Alarmed by the numbers of people losing insurance, some states are taking a more proactive approach to notifying and educating people about the new verification process for maintaining coverage. But a one-size-fits-all approach will not be effective in communicating with this diverse audience. Connecting with these different populations requires understanding their unique needs and preferences and delivering culturally sensitive content in multiple languages. Digital health solutions are well positioned to help states and providers achieve their shared goal of engaging Medicaid patients.

Combining digital communications with human efforts is critical to achieving this daunting task. Trust plays a role as well, as more people with Medicaid express wariness about their providers. The report “A Two-Way Street: Building Trust Between People with Medicaid and Primary Care Doctors” published by Public Agenda found that four in 10 say doctors need to earn their trust. Communicating with these patients in culturally sensitive and health literate language should be central to the strategy for engaging them to play a more proactive role in their healthcare.

By leveraging patient data and insights, technology can help personalize the content and optimize the outreach by channel, ultimately improving effectiveness and ensuring that patients do not get lost, while building a greater bond of trust between them and their providers.

As states continue to unwind the Medicaid continuous enrollment provision, there are opportunities to promote continuity of coverage among enrollees who remain eligible by implementing a patient engagement strategy that leverages digital communications along with human efforts to reach, educate, and activate patients.

Readers Write: The Illusion of EHR Interoperability

The Illusion of EHR Interoperability
By Pawan Jindal, MBBS

Pawan Jindal, MBBS, MHI is CEO of Darena Solutions of Chesterfield, MO.

Isn’t EHR interoperability great?

It would be, but there is a huge gap between the published standards and the reality. Sharing data among healthcare providers, health plans, and patients was supposed to be much easier now that EHR interoperability through FHIR-enabled apps is the universal standard. Developers should have been able to integrate their SMART on FHIR apps with virtually any EHR and have the resulting integrations work seamlessly across multiple platforms. Sadly, it is not working as intended.

In Q1 2023, the Office of the National Coordinator for Health Information Technology (ONC) reported that 95% of certified health IT developers met the December 31, 2022 compliance deadline to enable access to information through application programming interfaces (APIs) “without special effort.” However, our experience with FHIR app developers, providers, and EHRs shows that true EHR integration remains elusive, despite ONC’s claims.

Out of the nearly 300 EHRs certified by ONC to be interoperable with FHIR-enabled apps, only a few allow developers to integrate apps with their EHRs. By enforcing the Cures Update requirement only on EHR vendors, ONC is not penalizing providers, the ones who seem to be refraining from information sharing. Out of the total 763 claims of information blocking filed so far with the ONC, 85% of the claims (646) are against providers. This problem is further exacerbated by the fact that provider education on the benefits of information sharing from ONC is severely lacking.

The Information Blocking provisions of the Cures Act currently only mandate making data available to patients upon request. The EHRs have geared up to allow providers to honor these requests. However, if you ask providers, they say, “No one is asking for it, or I send them to the patient portal, or I ask them to fill out a request form to obtain a hard copy of their records.”

Most providers aren’t aware of the requirement to provide data to patients in an app of their choice. If healthcare is ever to achieve a reality that includes easily integrated apps facilitating the seamless sharing of patient data between organizations, it must actively engage providers in information sharing.

Healthcare has been working toward interoperability for a while through the creation of rules and standards. It’s been three years since the Centers for Medicare and Medicaid Services (CMS) adopted the interoperability rule, removing many barriers that prevented patients from accessing their health data. The rule also issued version 1 of the US Core Data for Interoperability (USCDI v1) standard that EHR vendors must meet for ONC certification.

An information blocking provision went into effect in 2021 requiring EHR vendors, providers and others to share the data specified in USCDI v1. That rule was expanded in 2022 to include even more types of data. Last year, ONC also published the Trusted Exchange Framework and Common Agreement (TEFCA), which sets a nationwide standard for interoperability and establishes the process for health information networks to become Qualified Health Information Networks (QHINs), a sort of “super network” for sharing data.

FHIR (Fast Healthcare Interoperability Resource) is the standard developed to enable this data exchange. It can be used on its own and with existing standards, like the USCDI and billing-related data elements used in EHRs. FHIR-based apps are designed to be used with any FHIR-capable EHR. It is important to note that the TEFCA agreement is meant to establish a minimum standard for performance across the healthcare continuum. Based on that, FHIR is on the map for future phases and is not required out of the box.

So why does widespread EHR interoperability remain an illusion despite ONC claims?

Glitches are to be expected any time there is development and adoption of a new technology standard, particularly one that must integrate with older EHR platforms. Companies, sometimes unwittingly, fail to disclose all the ins and outs of their products and capabilities. Take for example, NextGen Healthcare’s agreement this summer to pay a $31 million fine to settle claims that the company misrepresented its software’s capabilities and paid users kickbacks for their endorsements. Similar cases have resulted in settlements with other EHR vendors, including EClinicalWorks, Practice Fusion, Greenway Health, and Modernizing Medicine.

Even when considering glitches and a few bad actors, it’s become obvious that ONC certification alone doesn’t necessarily guarantee successful app integration in the field because developers, EHR vendors, and healthcare systems continue to struggle to achieve interoperability.

For its Health IT Certification Program, the ONC includes a Real World Testing annual requirement. According to the website, “The purpose of this Condition and Maintenance of Certification requirement is for Certified Health IT Developers to demonstrate interoperability and functionality of their certified health IT in real world settings and scenarios, rather than in a controlled test environment with an ONC-Authorized Testing Lab.”

Anyone with experience in IT development (or any complicated technology, for that matter) knows that what works well in the lab can fail in the field. That’s because real-world conditions and demands can be more challenging than what designers anticipated. This highlights the need for more realistic real-world testing from the ONC in addition to tests conducted by independent entities. Currently, each EHR tests its own application in the field. Unsurprisingly, they all seem to replicate the certification testing. We need a Consumers Reports-style impartial review for health IT.

In the meantime, app developers and other stakeholders can work with third-party experts who can guarantee EHR integration.

Readers Write: From EHRs to EOM: Enhancing Oncology Model Highlights Limitations of Current Clinician-Facing Tech

From EHRs to EOM: Enhancing Oncology Model Highlights Limitations of Current Clinician-Facing Tech
By Kathy Dalton Ford

Kathy Dalton Ford is chief product and strategy officer at Ronin of San Mateo, CA.

For years, healthcare leaders have prioritized improving patient access and care delivery through value-based care (VBC) initiatives. However, according to a 2022 report, value-based contracts only accounted for 7% of medical revenue among primary care specialties, 6% among surgical specialties, and 15% among non-surgical specialties. These percentages indicate that despite the efforts of healthcare leaders, there is still a long way to go to implement VBC programs fully.

To address this issue, the Centers for Medicare & Medicaid Services (CMS) introduced a range of value-based care models, including the Enhancing Oncology Model (EOM). EOM, a voluntary five-year model that commenced on July 1, 2023, aims to improve the quality of care while reducing costs through payment incentives and required participant redesign activities.

Provider organizations must use certified Electronic Health Record (EHR) technology as part of the required redesign activities. EHRs are ubiquitous, with nearly four in five office-based physicians (78%) and almost all non-federal acute care hospitals (96%) adopting a certified EHR as of 2021. However, EHRs facilitate billing rather than inform care decisions, lacking the all-important ePROs and daily insights into patient conditions to inform effective cancer care. 

While EHRs support billing and reimbursement, they present several challenges for physicians in delivering timely, quality patient care, resulting from time-consuming data entry, interoperability issues, un-optimized user interface design, and lack of standardization. These problems make it challenging to access vital patient information at the point of care, increasing the time required to document patient encounters and potentially leading to errors or missed details.

Many organizations don’t have the tools to implement VBC-based programs and payment models, making EOM’s implementation governance and reimbursement support critical in realizing these life-saving initiatives. Meeting EOM requirements cannot solely be fulfilled by care teams and EHRs alone. Health systems must adopt clinical decision-support technologies that consider the patient experience outside the hospital, connect patients to their care team, and integrate safe and ethical artificial intelligence (AI) to fill the gaps in existing capabilities and realize the benefits of value-based care.

Today’s AI technology can pull data from unstructured clinician notes, accelerate time-consuming chart reviews, and improve care by analyzing data to produce actionable predictive insights. By pairing AI with a robust decision support platform and ePROs, cancer centers can provide patients with 24/7 access to care teams, streamline patient-to-care team communications, engage patients, screen for social needs, deliver health education, and identify patients at risk for adverse events.

Health systems must adopt solutions incorporating safe and ethical AI tools that accelerate precise clinical care decisions and rise above the competition to leverage EOM and capture new revenue without the burden of adding more steps to their workflows. By doing so, healthcare leaders can improve patient access and care delivery while reducing clinical and administrative burdens and realizing the full benefits of VBC programs.

Ultimately, the goal of EOM is for patients to feel better supported in their care; have a clearer understanding of their diagnosis, prognosis, and outcomes; and adhere to their treatment plan. However, the tools and data to help clinicians meaningfully facilitate their job have yet to be available.

Hospitals now have an opportunity to leverage technology to help them realize the vision of comprehensive, coordinated cancer care.

Readers Write: Navigating the Future of Clinical Information Post-Public Health Emergency

Navigating the Future of Clinical Information Post-Public Health Emergency
By Greg Samios

Greg Samios, MBA is president and CEO of the clinical effectiveness business of Wolters Kluwer Health.

The global health community has acknowledged the official end of the COVID-19 public health emergency (PHE). The impact of the PHE had both positives (telehealth) and negatives. Of the latter, there are many, but the rampant flow of inaccurate and misleading information, also known as the infodemic, is a key one because of its direct impact on patients.

This infodemic highlighted the important role clinical decision support (CDS) tools play in distributing a high volume of reliable, trustworthy, and ever-evolving information to clinicians around the world. As the global health community looks to the future, there are a few lessons from COVID-19 to consider about the power of CDS.

The PHE may have ended, but what about the Infodemic?

The end of the PHE offers a moment of reflection about where the industry goes next to ensure that CDS continues to support clinicians through distributing reliable, transparent, and consistent information across care teams. It also opens the possibility for a period of uncertainty and unpredictable increases in the variability of care.

Looking ahead – as healthcare leaders consider both ongoing threats of COVID-19 as well as the inevitable next pandemic – CDS resources could be leveraged to improve speed and transparency and more effectively reach public health goals during an infodemic. The industry needs to focus on how it can be agile in distributing continually emerging and changing information in the next PHE.

But there is another challenge looming to further deepen the entrenchment of the infodemic: the arrival of generative AI, including ChatGPT. While generative AI offers potential for healthcare, it may also present risks if not developed and applied responsibly. This could be particularly critical around its use in clinical care.

CDS everywhere, including virtual

The PHE helped deliver new avenues for patients to interact with healthcare providers, such as virtual visits. It also proved and elevated the important role that local retail pharmacists play as an extension of a patient’s care team – providing COVID tests, vaccines, treatment, and counsel to patients, among other key responsibilities. The challenge for the future will be to ensure that no matter where patients interact with their care team, they receive the most optimal and consistent care as possible.

In tandem with these shifts, it’s crucial that healthcare systems work together and provide smart, consistent, and accurate information. CDS resources offer a standard approach to align the thousands of micro-decisions clinicians make every day, from physicians in the emergency room to primary care doctors at urgent care to virtual care at home and pharmacists at the neighborhood pharmacy.

Closing the care variability gap

There is still a great deal of care variability, depending on which clinician a patient visits, where the patient lives, how much insurance and social support a patient has, and numerous other factors. Regardless of circumstances, clinicians should still have access to the most recent data and treatment recommendations. COVID-19 demonstrated that when information is widely shared, CDS resources can swiftly close the gap whether clinicians are eight or 8,000 miles apart.

More data, more insight

CDS is standard for clinicians to search data to diagnose patients. But the power of those searches can also create new data that can provide a broader set of insights. By analyzing clinician search queries, CDS enables providers to see around corners and proactively observe trends and understand usage patterns, such as which clinical questions are most important.

CDS resources can also share new medical updates with millions of providers and push notifications within the workflow of electronic medical record (EMR) systems to quickly educate clinicians with treatment recommendations that are trustworthy, verified, and improve patient outcomes, which can be incredibly valuable during a public health emergency.

Ultimately, it’s to everyone’s benefit to create an ecosystem where clinical knowledge systems and EMR vendors can work harmoniously to capture and inform point-of-care decisions.

During the PHE, global healthcare leaders learned how to adapt and make changes to everyday healthcare operations to improve patient outcomes. To make progress as an industry towards closing the care variability gap, and to ensure we are prepared for the next PHE, health organizations should seek a CDS partner that can provide both access to trustworthy and timely information, continuity to support patients no matter where they seek care, and provide insights to benefit the entire healthcare system.

Readers Write: Leveraging a Digital Ecosystem to Simplify Specialty Medication Onboarding

Leveraging a Digital Ecosystem to Simplify Specialty Medication Onboarding
By Julia Regan

Julia Regan, MBA is founder and CEO of RxLightning of New Albany, IN.

Specialty medication onboarding has historically been a manual and frustrating process, riddled with complexities, administrative hurdles, and delays. However, the convergence of technology, interoperability, and a robust ecosystem of partners can revolutionize this experience, offering a glimmer of hope and path to a smoother, more efficient onboarding journey for patients.

The Complexities of Specialty Medication Onboarding

Getting a patient started on a specialty medication goes well beyond the enrollment forms. In most cases, onboarding a patient requires enrollment paperwork, benefit verification, prior authorization, financial assistance, and ongoing communication between multiple parties. The trouble is that each of these steps has typically occurred in a standalone manner, without any connection to the other steps in the process. This creates an opaque and disjointed journey for patients and providers, slowing time-to-therapy and causing avoidable administrative burden.

In addition to the process being fragmented and unclear, some parts of the journey are still completed using paper, phone, and fax – hindering transparency from the start! Whether completing the initial enrollment, approving a PA, or submitting Patient Financial Assistance forms, we are living in a world in which the system we’ve created impedes patient outcomes – instead of improving them. We must do better.

The Power of Technology and Collaborative Ecosystems

Thankfully, the advent of interoperable technology and collaborative ecosystems are beginning to bring about significant improvements to the specialty medication onboarding experience.

The first step in creating a better onboarding journey is eliminating the need for paper-based forms and communication. By leveraging fully digital portals for documentation submission and collection, we can provide immediate feedback to users about missing information, statuses, and next steps. And once digital becomes the norm, providing transparency to key stakeholders is no longer an impossibility.

The next – and arguably most important – step is developing an open ecosystem, where each participant plays a vital role in the medication onboarding journey. Each interconnected partner must be aligned in achieving a shared vision, and each plays a critical role in the final delivery and adherence of the medication. Providers and pharmacies can review and confirm coverage information, care teams can find and submit financial assistance applications, and patients can be more effectively supported by manufacturer and hub support teams.

Stakeholders should not need to log into multiple systems to manage one patient journey; instead, they should have access to one platform with all the data and integrations they need. With a truly connected ecosystem, each stakeholder can make informed decisions based on accurate and up-to-date information, ensuring the timely initiation of therapy without unnecessary hurdles.

What’s Next for Specialty Medication?

While today’s specialty medication landscape is already complex, tomorrow’s is set to become even more convoluted. At the 2023 Academy of Managed Care Pharmacy meeting, IPD Analytics shared that nearly 80% of the drugs the FDA is expected to approve in 2023 are specialty drugs, up from 68% in 2020.

How the industry navigates this wave of specialty drug approvals could significantly influence patient care. From my perspective, a transparent, interoperable system could address many of the previously mentioned challenges by streamlining communication and providing real-time access to critical information that can be used to support patient engagement, affordability, and adherence.

As specialty pharmacy continues to expand and evolve, the need for a unified, comprehensive medication onboarding ecosystem becomes increasingly important. By harnessing the power of technology, interoperability, and a collaborative ecosystem, we have the opportunity to revolutionize this space. Together, we can build a world in which every stakeholder, from the provider to the patient, is empowered to navigate the intricacies of the specialty onboarding experience.

Readers Write: The Shift Toward an Employer-Driven Market in Healthcare Technology

The Shift Toward an Employer-Driven Market in Healthcare Technology
By Mike Silverstein

Mike Silverstein is managing partner of the healthcare IT and life sciences practice of Direct Recruiters, Inc. of Solon, OH.

Throughout the last few years, the healthcare technology market has been largely candidate driven. During the “grow at all costs” period that started in late 2020 and continued through fall of 2022, capital was cheap, and companies were doubling down on product that was to be delivered by expensive and newly hired engineering and data talent.

That has slowed tremendously in recent months and has been coupled with significant layoffs across most of big and medium tech. What felt like an inelastic demand for technical talent over the last number of years, which corresponded to growing compensation demands, has flipped.

Below are three market trends we are seeing that signify a shift toward an employer-driven market.

Technological Advancements

It’s no secret that technology continues to change the landscape of the global workforce. The continuous stream of new AI and automation tools being introduced has the potential to change processes, procedures, and potentially even replace human labor in some situations. According to a March 2023 report from Goldman Sachs, the automation of certain tasks could disrupt a staggering 300 million jobs worldwide.

As these changes evolve, employers’ expectations of current and prospective talent are to be able to adapt and leverage new technology to their advantage versus letting it replace them.

An Emphasis on Talent That Has a Near-Term ROI

Sales, business development, demand generation marketing, customer success, and FP & A roles are crucial right now. Investors are demanding greater discipline from their portfolio companies as the cost of capital has increased and the bottom has fallen out of company valuations, particularly in tech. Right now, each company is tasked with showing a path to break even and/or profitability. No investor wants their portfolio company to have to go out for a fundraise right now for fear of a down round.

While healthcare technology employers hire and retain talent, the pressure is high for candidates to showcase that they are results driven to land great roles. If you can help make a dollar, protect a dollar, or count a dollar – sales and marketing, customer success, and accounting and finance, respectively — there are still strong opportunities in the market. As companies strive toward profitability in a tumultuous time, there may be more uncertainty for roles further away from revenue.

The Abundance of Tech Talent

With recent layoffs, there is now an abundance of healthcare technology talent on the street with far less demand for its services. As a result, passive candidates have become more risk averse. Clients are realizing there is a bit of an opportunity to buy low(er) on some needle-moving talent.

A lot of mediocre candidates did really well for themselves over the last couple of years. Healthcare technology companies are seeing an opportunity to top grade on positions where they settled in the last 24 months, and there is added scrutiny on every candidate in the hiring pipeline.

As we face this potential shift in the market, talent that has a track record of being able to perform and execute in a capital-constrained environment will continue to thrive. Candidates who are more entrepreneurial in the traditional sense — in that they are comfortable doing more with less, versus relying on the ability to obtain unlimited growth capital whenever needed — are still in high demand, along with those with a strong accounting and operations acumen.

Readers Write: Automation’s Role in Patient Engagement Strategies

Automation’s Role in Patient Engagement Strategies
By Lew Parker

Lew Parker, MSIS, MBA is CTO of Arrive Health of Denver, CO.

Thanks to ChatGPT, there’s no shortage of content on the topic of AI and automation’s impact on the future of the workplace. You don’t have to be a technology professional to understand that we’re approaching — if not already in — a pivotal moment that will change how we all do work, much like how the IPhone changed our everyday lives.

While AI may feel like a daunting new topic, the reality is that automated solutions have been widely adopted within all industries for quite some time. Healthcare is no exception, with AI-driven tools commonly being used to make the revenue cycle more efficient.

A prime opportunity within healthcare is bringing AI-driven solutions from the back office and using them to create better patient experiences and outcomes. If we can leverage AI-driven automation tools to drive engagement throughout the patient journey while also alleviating administrative burden for overworked providers, why wouldn’t we invest in tools that do that? The key is determining when and where within patient engagement strategies to implement automated approaches, and how to do so in a thoughtful way that preserves data security and protects patient PHI.

Medication adherence presents an ideal opportunity to enter the waters. According to the World Health Organization, medication adherence can directly impact patient outcomes more than a specific treatment itself. But traditional methods to drive adherence — including relying on staff to conduct massive amounts of personalized outreach once a script has been selected – leave room for improvement, as evidenced by one in five prescriptions never being filled and burned out pharmacists, providers, and care teams.

One factor contributing to non-adherence is the engagement gap that exists from the time a provider writes a prescription to when the patient arrives at the pharmacy counter. A lot can occur during that period to prevent a patient from picking up the prescription. AI-driven solutions, specifically patient outreach tools to connect with patients during this time in a more natural and ultimately human way. Anyone who has worked with traditional chatbots or robocalls can attest to the frustration of both patients and implementers when the conversation strays outside the boundaries of what computers could handle. That has fundamentally changed, and now we can bridge the gap to support patient engagement and medication adherence by:

• Reducing significant volumes of manual and repetitive tasks by automatically sending information over text messaging about topics such as onboarding, medication information, and refill reminders.
• Creating additional patient touch points along the medication adherence journey while reducing the overall cost of outreach.U
• sing AI to identify which patients have barriers to adherence and then prioritizing them for personalized pharmacist outreach. This will allow pharmacy staff to focus on high-value, high-touch care.

It’s impossible to expect pharmacists and providers to conduct the level of personalized outreach needed across all populations served without the help of technology. There are simply not enough hours in the day. However, using AI-driven tools as described above will drive better patient experiences, engagement, and outcomes while also helping an already overburdened healthcare workforce. As IT leaders, it’s our job to encourage innovation while also supporting responsible and impactful use of AI in new ways, including bringing these tools to patient engagement strategies.

Readers Write: Cutting Through the Hype: Navigating AI in Healthcare

Cutting Through the Hype: Navigating AI in Healthcare
By Michael Burke

Michael Burke, MBA is founder and CEO of Copient Health of Atlanta, GA.

First, a confession. Our company leverages machine learning in our operating room utilization software solution. As such, we stand to benefit from the AI hype machine that is running at full speed. But I promise you, the intent of this paper is not self-promotion; it’s to help you distinguish true AI value from mere marketing hype.

Understanding the categories of AI, and a little about how each tool set works, is essential. However, wading into the weeds a bit is an unfortunate requirement of doing this.

The Basics of AI: What Healthcare Executives Should Know

Before delving into the nuances of true versus misleading AI claims, let’s first understand some fundamental AI categories relevant to healthcare. The items below aren’t a comprehensive list, but they do capture some of the most common and important categories you’ll see:

Machine Learning (ML)

Machine learning is a branch of AI that uses specific algorithms to analyze and learn from large amounts of data. This “training” process, and subsequent testing, results in a model that can make predictions or decisions without being explicitly programmed to perform the task.

Training methods of different ML algorithms include:

• Supervised learning. Supervised learning is like teaching a child with the help of a teacher. The teacher knows the correct answers and provides guidance, and the child learns from this guidance. An example of supervised learning in healthcare could be predicting whether a tumor is malignant or benign based on a set of labeled tumor data. In this case, the “teacher” is a labeled dataset where each tumor is classified as either malignant or benign. The ML model is trained on this data and can then predict whether a new, unlabeled tumor is malignant or benign based on what it has learned.
• Unsupervised learning. Unsupervised learning is like a child learning through exploration without guidance from a teacher. The child learns about the world by observing and interacting with their environment. An example of unsupervised learning in healthcare could be patient segmentation, where healthcare providers group patients into different categories based on their health data. There are no pre-labeled categories here; the ML model must learn to identify patterns and structure in the patient data to determine how best to group the patients. For instance, an unsupervised learning algorithm could analyze patient data to identify clusters of patients with similar health characteristics, which might correspond to different risk groups or disease subtypes.
• Semi-supervised learning. Semi-supervised learning is a combination of supervised and unsupervised learning. The model learns from a combination of labeled and unlabeled data. A semi-supervised learning example could be if a hospital had a large amount of patient data, but only a small portion of the data is labeled. Perhaps only a small set of the patients’ records include a diagnosis for a certain disease. The ML model can use the small amount of labeled data to learn about the characteristics of patients with that disease and then apply this learning to the large amount of unlabeled data to predict which of those patients might have the disease.
• Reinforcement learning. Reinforcement learning is more like teaching a dog to perform a new trick. You don’t tell the dog explicitly what to do. Instead, the dog tries different actions, and you give it a treat (a reward) when it performs the action you want, like sitting or rolling over. Over time, the dog learns which actions will earn it a treat and starts performing those actions more frequently. In a healthcare context, consider a machine learning model that is trying to optimize treatment plans for patients with chronic conditions. To compare to our dog training explanation, the ML model is the dog, the different treatment plans are the actions the dog tries, and the patient health outcomes are the rewards. The model tries different treatment plans (actions) and observes improvements in patient health (the reward). Over time, the ML model learns which treatment plans lead to better patient health outcomes, much like the dog learns which actions earn it treats.

ML excels at tasks where patterns and structures can be discerned from data, such as prediction (predicting hospital readmission rates based on patient data), classification (classifying skin lesions as benign or malignant based on image data), and clustering (segmenting patients into different risk groups based on their health data).

Deep Learning (DL)

Think of deep learning as a team of detectives working on a case. Each detective looks at a part of the evidence and makes their own observations. They pass their findings to a senior detective, who then makes more complex observations based on the initial detectives’ findings. This goes on until the chief detective (the final layer of the network) makes a decision based on all these observations.

In the world of AI, each ‘detective’ is a layer in an artificial neural network. Each layer looks at some aspect of the data and passes on its findings to the next layer. This allows the network to learn from simple features at lower layers to more complex features at higher layers.

Let’s translate this into a healthcare example. Consider a deep learning model analyzing an MRI scan to detect a tumor. The initial layers might look for basic features like edges or colors. The next layers might recognize more complex patterns like shapes or textures. And the higher layers might identify the specific features of a tumor. Just like our detective team, each layer contributes to the final decision, allowing the model to accurately identify whether a tumor is present.

Deep learning excels at tasks involving unstructured data such as images, audio, and text. For instance, deep learning algorithms can analyze MRI images to detect tumors, listen to a patient’s speech to diagnose mental health conditions or analyze electronic health records to predict patient outcomes.

Natural Language Processing (NLP)

Natural language processing (NLP) involves the interaction between computers and human (natural) languages. This technology allows computers to understand, interpret, and generate human language in a valuable way. At its core, NLP involves machine learning to automatically learn rules by analyzing a set of examples and making a decision based on them. This decision could be understanding sentiment, translating languages, or converting speech to text.

In the healthcare sector, NLP can be used to interpret clinical documentation, analyze patient feedback, or enable natural language user interfaces (e.g., chatbots for patient engagement).

Generative AI

Generative AI involves creating new, previously unseen content. Think of it like an AI artist that creates new works based on styles it has learned from. Generative AI is not limited to any particular type of content and can generate images, text, music, and other types of data.

In healthcare, generative AI could be used to create synthetic patient data that can be used for research or training purposes without compromising patient privacy. For instance, a generative model could be trained on real patient data and then generate new data that maintains the statistical properties of the original data (like the distribution of different diseases or the average patient age) but does not correspond to any real individual patient. This synthetic data can then be freely used without worrying about privacy violations.

Computer Vision

Computer vision is like teaching a computer to ‘see’ and interpret visual data in the way humans do. This technology is extremely versatile, being used in everything from self-driving cars to facial recognition software.

In healthcare, computer vision is often used in medical imaging to detect diseases and conditions. For example, computer vision algorithms can be used to analyze X-rays, MRIs, or CT scans to detect tumors, fractures, or other abnormalities. It’s also used in telemedicine solutions, where computer vision algorithms can help monitor patients and detect abnormalities or changes in their condition. Computer vision is also used in robotics.

Knowledge Representation and Reasoning

Imagine AI as a detective solving a complex case. To do this, it needs  a vast amount of knowledge about the world, along with the ability to reason with this knowledge to draw conclusions. That’s what knowledge representation and reasoning AI do.

In the healthcare domain, such AI can be used in clinical decision support systems to aid physicians in diagnosing diseases. The AI system has access to a vast amount of medical knowledge and can reason with this knowledge to provide suggestions to physicians.

It’s hard to decide where to draw the lines when categorizing types of AI. For instance, that the problem of tumor detection often involves computer vision as part of a series of machine learning models that feed together into a deep learning network. Additionally, a field like robotics is sometimes considered its own AI category and other times considered an application that uses specific categories of AI. I’m confident that people smarter than me will sometimes disagree with my categorization choices.

Identifying Genuine AI: A Guide to Avoiding the Hype

While understanding AI categories is a good starting point, the key to discerning genuine AI applications in healthcare software lies in recognizing when these technologies add real value to a process or outcome. And remember: AI, at its core, should aid decision-making, not replace it.

It’s easy for marketing campaigns to dress up their solutions with the AI label, but there are several ways in which the reality may fall short of the hype.

The AI Imposter: Recognizing Automation Dressed Up as AI

AI, including ML, DL, and NLP, learns and improves from data over time, enabling complex decision-making that is generalizable and extends beyond predefined rules. In contrast, rules-based automation, though beneficial in certain contexts, lacks this level of complexity and adaptability.

Consider, for example, a software solution that sends alerts when patient vitals reach certain thresholds. This represents a rules-based automation system, not AI. A genuine AI solution might continuously analyze patient data, learn from it, predict potential health risks before they become critical, and even suggest personalized treatment plans. This shouldn’t imply that a solution that leverages AI is necessarily better than a rules-based automation solution. However, beware of vendors dressing up automation as AI to take advantage of the hype as a rationale to increase price.

The Overkill: Unnecessary AI Implementations

Some solutions may incorporate AI where it’s unnecessary, serving more as a marketing tool than a feature that adds value to the end user. An example could be an element of a software solution that matches tasks to individuals based on skills. It might use an unnecessarily complex ML categorization algorithm where a simple lookup table would have sufficed.

The Overstatement: Claiming Unrealistic AI Capabilities

Some vendors may oversell what their AI can achieve. While AI can indeed help predict patient outcomes, claiming perfect accuracy is unrealistic and potentially misleading. Also, prediction accuracy in machine learning is a moving target as training data sets change over time, sometimes in response to the ML-based intervention itself.

Downplaying the need for quality data can lead to overstating anticipated results. AI systems are only as good as the data they’re trained on. If a solution downplays the importance of data quality, quantity, or diversity, be skeptical.

The Overlook: Ignoring the Need for Human Oversight

True AI applications in healthcare are designed to support and enhance human decision-making, not replace it. If a solution suggests that its AI can replace human judgment entirely, it’s likely overhyped. It could also be dangerous.

Conclusion

While Copient Health indeed benefits from the AI boom, we urge discernment when it comes to AI in healthcare. Understanding the basics and recognizing when AI genuinely adds value is critical. The future of healthcare is undeniably intertwined with AI. With a robust grasp of the subject, you’ll be primed to guide your organization into a more efficient, patient-focused era.

Readers Write: It’s Time for EHRs to Alleviate, Not Exacerbate, Clinician Burnout

It’s Time for EHRs to Alleviate, Not Exacerbate, Clinician Burnout
By Nancy Pratt, RN

Nancy Pratt, RN, MSN is senior vice president of clinical product development of CliniComp of San Diego, CA.

We all see the headlines – clinician burnout is taking an enormous toll. It is estimated that 47% of US healthcare workers may leave the profession by 2025. One of the most-cited sources contributing to burnout is electronic health records (EHRs), with nearly 60% of physicians in one poll suggesting that EHRs need a complete overhaul.

In my work with clinicians, their biggest frustration with EHRs is time. So much of their time is spent doing manual, EHR-related tasks, causing them to spend nearly twice as much time in the EHR as they do with patients. A poll by Stanford Medicine found that hospital-based physicians spent 25 of the 37 minutes on behalf of each patient in the EHR.

It doesn’t need to be this way. By collaborating with physicians, nurses, pharmacists, and other clinicians, EHRs can become a trusted part of care delivery processes, freeing clinicians to focus on their patients and recapture the most satisfying qualities of their professions. With a focus on human-centered design, EHRs can help reduce clinician burnout in three ways:

Capture documentation naturally as part of the workflow

Given the frustration with documentation, EHRs need to move beyond focusing on transactions to creating efficient, supportive workflows for all clinicians. One process that is often frustrating and burdensome is medication reconciliation. A well-designed EHR workflow can present this critical step at a natural point in the care process. By presenting a side-by-side comparison of what’s current and what’s needed for the patient, the process becomes a fast and accurate way for physicians to manage medications.

Bring device data into the workflow automatically

Using standard integration protocols, such as APIs, EHRs can automatically integrate data from bedside monitoring and other devices into user-focused workflows. Instead of requiring nurses to enter data manually, the EHR should be fully integrated, perform calculations automatically, and present in a user-friendly way. In addition to reducing errors and manual tasks, nurses working with monitored patients report saving as much as 15 to 20 minutes per patient per shift with this level of integration.

Remove downtime as a barrier

It’s surprising that system maintenance and upgrades still require planned downtime for many EHRs, slowing down care delivery with manual, paper-based workarounds. Unplanned downtime, experienced by 96% of organizations according to one study, can be even more onerous. EHRs should be built upon modern foundational architectures that don’t require scheduled downtime and have built-in redundancy to prevent unplanned downtime. Preventing downtime alleviates a common source of stress for clinicians when care delivery is hampered by lack of access to patient information.

It’s time for the industry to listen when so many clinicians say EHRs need to be revamped. Using flexible, well architected technologies and collaborating with clinicians, EHRs can enhance rather than hinder care delivery. At last, EHRs can support clinician wellness by enabling them to provide the highest quality patient care, bringing the joy of helping patients back into their day-to-day work.

Readers Write: Return Data to Hospitals and Researchers for Patients

Return Data to Hospitals and Researchers for Patients
By Amanda Borens

Amanda Borens, MS is chief data officer of Aridhia Informatics of Glasgow, Scotland.

As a data scientist formerly working in global health, I have firsthand knowledge of how challenging it is to recruit patients for clinical trials and observational studies, and how transformational a culture of open collaboration and data sharing can be. I have cancer diagnoses in my history, accompanied by so much fatigue from being extensively examined and poked and prodded, but I’m also a scientist in the health technology industry with appreciation for the way that humans advance medical knowledge.

I have eagerly signed up for clinical studies and taken on extra burdens, such as blood draws and filling out forms, to help researchers as well as those in training. Even when that means being examined by a nervous medical student and then repeating myself to a physician.

Then why did I balk when my health system asked me to participate by openly sharing my EHR data for research?

I collaborated with renowned medical ethicists who taught me to embrace the idea that a patient owns their data. That idea was cemented in me when I became a participant in various studies that led to peer-reviewed publication. I had been working in my non-profit world to aggregate data from multiple sources to learn more about rare diseases. I became frustrated that companies would keep shareable data confidential to avoid letting competition find insights that they had missed. I was shocked to learn that many scientists in academic settings hoard data in the same way, fearing that someone might find something they missed in the data and publish it.

In all cases, the stewards of the data seemed to forget that patients sacrificed time, blood, energy, and more to help all humanity, not the personal careers of their physicians or the bottom lines of sponsoring drug companies. That intention matters, and is worth honoring.

This brings me to my personal hesitation when asked to share my EHR data for research. I’ve been to data science conferences where abstracts were presented by employees of this health system. I know the kinds of questions my data would answer. Those questions tend to be focused on how to keep hospitals profitable. While I know that this is a valid concern for administrators in hospitals, I want my personal health information to be used to help other patients like me have a better experience, and I don’t much care what that would cost.

I want to be a piece of data that led to a cure for this, a better treatment for that, an earlier detection of my cancer for others, a less-invasive surveillance journey, or a better experience with caregivers. I want to share my EHR data with my incredible team of oncologists and researchers so they can learn. I want my data to be compiled with that of others so that they can learn more and faster.

However, I know that my doctor won’t be able to access aggregated data that way in my clinic. I know he had to use an Excel spreadsheet to keep track of data in the landmark oncology study he recruited me to join.

Conversely, aggregated claims databases can be used to answer questions about health economics. Some cost money to access, and sometimes researchers can access those external data sources with cost waived. Beyond payer-aggregated data sets, Epic has a respectable database and Cerner does too, both of which are valuable for review of de-identified patient data. But what about empowering researchers and clinicians in their unique patient communities? Shouldn’t we be honoring the patients’ commitment to advancing medical science by empowering clinicians and researchers to more easily use more data in their own hospitals? What does that look like?

The US government mandates that healthcare IT developers like Cerner and Epic provide their customer base with a certified FHIR API to support patient access to health information by December 31, 2022 as part of the 21st Century Cures Act Final Rule. This mandate requires that certified health IT developers publish “service base URLs” or “endpoints” for all customers in a machine-readable format at no charge. I hope this will be an inflection point.

Additionally, patient-focused drug development mandates are demanding greater listening to the voice of the patient, and that means tighter connection between pharmaceutical companies and real-world data from clinical settings. Hospitals with an investment in a next-generation research environment will be able to procure industry funding to collaborate in a secure, audited cloud environment that is dynamic and connected to anonymized EHR data alongside observational or interventional study data. A nice side benefit? That same (already funded) environment may provide a subset of hospital researchers with identifiable patient data that can be used to implement research findings into clinical practice in a timelier fashion.

As a patient, I’d donate my data to that hospital in a heartbeat.

Imagine a world where patients and clinicians collaborate to improve healthcare, then take a look at what Great Ormond Street Hospital for Children is doing with their research environment for a wonderful example. Pediatricians there have been studying precision dosing regimens and collaborating across continents to share dosing models where pediatric populations were excluded from clinical trials. This isn’t happening in one hospital’s data warehouse, and it isn’t happening with access to a single EHR or aggregated repository. It’s possible because of connecting different types of data of across borders and across time, but all in a next-generation research environment that connects people.

The transformative possibilities do not end there. What if a research hospital was able to collect clinical data from EHR and combine it with multiomics data from an academic research university where bioinformatics pipelines are optimized to provide a list of variants for each patient? How might they learn about patient subpopulations and disease progression or predict responses to interventions? That sort of collaboration should be the norm, but it requires us to think bigger than one data warehouse, one data type, or one organization at a time. Let’s make it happen.

Readers Write: Bringing the Call Center Into the 21st Century

Bringing the Call Center Into the 21st Century
By Ben Moore

Ben Moore is chief product officer of PerfectServe of Knoxville, TN.

Like some of you, I was compelled to dip my toe in the healthcare waters because of a personal experience.  It was over a decade ago, but I still remember it like it was yesterday.

My wife was pregnant with our daughter, who came weeks before her due date. Both my wife and daughter faced serious complications, and my stress level was exacerbated significantly by care delays that seemed almost entirely caused by inefficient communication. Simple questions and follow-ups that should have taken minutes ended up taking hours. These delays put both my wife and daughter’s health and safety at risk. 

For example, to reach the specialist, a nurse would have to call the hospital operator, who then had to manually track down and page the correct specialist on call, which would often start a process of telephone tag among nurses, operators, and physicians. The physicians were, and in some cases still are, carrying one-way pager devices, which would cause lots of disjointed one-way communications.

At that time, we had purchased the first-generation 3G-connected iPad for my wife while she was in the hospital. She was able to instantly text friends and family from her hospital bed and engage in social media. I was perplexed. Why weren’t the care teams in the hospital equipped with similar state-of-the-art communication tools? My wife was able to engage in real-time and media-rich communication, but hospital staff were limited to 140-character alphanumeric pager messages, telephone callbacks, and voicemails. The contrast was stark, and to be frank, seemed a bit ridiculous.

I can tell you that archaic technology and processes like the ones I described above are still prevalent across healthcare today. As just one example, even though over 70% of communications into and out of a hospital run through the call center, these antiquated tools are still, in many cases, a major part of directing that very important traffic. That’s alarming.

By upgrading technology that has its roots in the 1990s, the call center can become a true hub for patient engagement and clinical collaboration. Here are four steps you can take to bring the call center into the 21st century. 

1. Head for the cloud. Eighty percent of call center systems still use on-premise technology. This is problematic, because you’re only going to reliably connect stakeholders across all sites of care with a cloud-based solution. I once witnessed a fan blow up on a hospital PC server, and it completely knocked out communication with incoming ambulances. With a cloud-based system, this kind of risk is gone. If your EHR is already in the cloud, your call center should be there, too.
2. Unify clinical and patient communication. Rather than forcing operators to play middleman, go with a system that allows providers to communicate directly with patients. This kind of setup can be achieved within a broader ecosystem that facilitates both clinician-to-clinician and patient-to-clinician communication, meaning all communication is initiated and captured in one platform. This means less complexity for providers, easier access for patients, and greater transparency — especially with EHR integration — for anyone who needs to reference communication history.
3. Integrate, integrate, integrate. Speaking of integration, I actually think it should be a bigger focus for the call center than things like analytics or standard performance metrics, which you find in most call centers today. A properly executed integration plan can reduce the manual labor of operators by over 80%. Key integrations should include the EHR, patient flow systems, CRM, and scheduling platforms to provide operators with a single pane of glass.
4. Embrace smartphones. You need a call center platform that embraces smartphones. The vast majority of physicians use them, and integrating them with the call center ecosystem allows for things like time-sensitive communication — say, a team alert initiated by an agent about an incoming hip fracture patient — that can actually be monitored to verify that recipients have received and read all necessary information. No more guesswork! And please, whatever you do, make sure to ditch the pagers. No agent wants the page-and-pray technique to be central to their everyday duties.

Clinical call centers — like hospital switchboards, patient transfer centers, and answering services — can have a tremendous impact on everything from care coordination to patient experience to health outcomes. After decades of neglect, it’s time to give them the modern infrastructure they deserve to unleash their true potential.

Readers Write: Should Health Systems Become Banks?

Should Health Systems Become Banks?
By David Stievater

David Stievater, MBA is a partner with CWH Advisors of Boston, MA.

Are healthcare organizations expected to now run banks in addition to delivering patient care services? To help patients pay for their care, providers have become lenders, offering patients the equivalent of unsecured, zero-interest loans backed by hospital balance sheets. It should come as no surprise that with additional financial stress lingering from the COVID pandemic, providers are looking for additional funding sources and financing alternatives for their patients.

Overall US healthcare spending hit 18% of GDP in 2021, up from 5% in 1960. Total patient out-of-pocket spending, not including insurance premiums, totaled $433 billion, according to the latest numbers from CMS. The out-of-pocket dollars owed by patients has risen at twice the rate of US GDP growth since 1960. More and more, patients find they do not have the cash on hand to cover medical expenses.

The financial challenges facing patients are well documented. In 2021, the Federal Reserve Board reported that 17% of adults had major, unexpected medical expenses in the prior 12 months, with the median amount between $1,000 and $1,999. The same research also said that 23% of adults went without medical care due to an inability to pay. These types of findings are corroborated by other reports, including Synchrony Financial’s Lifetime of Healthcare Costs study in 2022 in which one in four individuals said they delayed a recommended procedure due to cost. The Synchrony study also indicated that when care was delayed due to cost, respondents ended up with additional medical issues 50% of the time.

Patients will benefit from more funding sources and options to reduce and eventually pay off balances owed to healthcare organizations.

Unlike other industries, providers generally bill patients after care is delivered. They carry the patient balance owed to them as a receivable, to be collected once insurance has calculated its portion. By billing after the fact, providers essentially make unsecured loans to patients.

Many decades ago, when the patient portion of provider revenues was relatively small, it was common for unpaid patient balances to simply be written off. It wasn’t worth the effort to pursue them. Providers did not tend to accumulate large unpaid patient balances on their books for long.

Over time, as the size and frequency of uncollected patient fees grew, provider organizations looked to new propensity-to-pay algorithms, early pay discounts, early out vendors, recourse financing, pre-service collection of co-pays, in-house payment plans, and other tactics to increase the yield on the amount owed by patients. Debt collectors were asked to pursue patients deemed able to pay. In some unfortunate cases, the most aggressive health systems have sued patients (including their own employees) to recover unpaid balances.

Provider executives say they work hard to balance compassion for patient circumstances with a desire to collect from patients who they believe can afford to pay at least some of what they owe. Additionally, they would prefer not to be in the lending game and tie up so much of their balance sheet with unpaid patient balances.

We estimate that 45-55% of a typical health system’s patient balance, after insurance (including full self-pay), is never collected and is converted to charity care, some other form of financial assistance, or written off as bad debt. Providers too will benefit from offering patients more options to fund their care and reduce and eventually pay off balances.

Our 2022 PatientPay study — which included 38 in-depth interviews with executives at health systems, hospitals, and large single/multi-specialty medical groups — indicated that providers will increasingly look outside their organization for solutions. One hospital CFO said that “a lot more people are going to finance their portion of what they owe,” Another revenue cycle executive said, “We are definitely going to provide more third-party financing options and less in-house over the next two years.” Overall, 61% of the executives said they expect to make greater use of third-party patient financing over the next 24 months.

Part of the impetus to look externally for payments solutions is that the COVID pandemic has made it harder for healthcare organizations to staff their revenue cycle operations. Still, much of the motivation to seek third-party assistance is simply the need to find more flexible payment and financing options for patients that free up the health system’s balance sheet.

The study noted that the biggest investments by large provider organizations are centered on creating a more retail-like payment experience for patients. COVID pushed more dollars here immediately in the form of contactless terminals and accelerated efforts around card-on-file and other portal-based technologies. Patients can expect improved communication via portal-driven emails, text to pay, and mobile applications.

Entrepreneurs are also putting their creative minds to work and generating new solutions to help providers match their patients with available financial assistance and philanthropic programs. Companies such as TailorMed/Vivor, Annexus Health (AssitPoint), RIP Medical Debt, and Atlas Health claim their approaches are generating meaningful ROI for providers.

Longer term, providers must focus on providing patients with an accurate estimate of their costs. This will enable them to collect the patient responsibility before care is delivered, without waiting for insurance adjudication and without turning their healthcare organization into a lender.

Readers Write: Rural Hospitals Need Federal Assistance to Strengthen IT Security Posture

Rural Hospitals Need Federal Assistance to Strengthen IT Security Posture
By Kate Pierce

Kate Pierce, MSMIITA is senior virtual information security officer and executive director of the Subsidy program of Fortified Health Security of Franklin, TN.

The majority of my career in healthcare IT has been dedicated to working for a small and rural hospital, leveraging technology advancements to improve patient care and to keep those systems safe from cyberattacks. I spent 21 years with North Country Hospital in Vermont, starting as a systems analyst and working my way up to chief information officer and chief information security officer.

Growing up in a rural community in northeast Vermont, I have a deep understanding of the challenges faced by smaller hospitals, especially those in rural settings. I also understand how vital these organizations are to the communities they serve.

When I was asked to testify before the US Senate’s Homeland Security and Governmental Affairs Committee on the challenges that small and rural hospitals face in managing an effective cybersecurity program, as well as barriers to adequate funding and human capital constraints, it was an honor to do so, as this is a topic near and dear to my heart.

Even without the cybersecurity challenges, rural hospitals are experiencing unprecedented staffing and budget constraints. More than 40% operate in the red, and nearly one in three is at risk of closure. When it’s a daily challenge to deal with basic healthcare delivery while managing higher labor costs and shrinking margins, cybersecurity isn’t a top priority for most hospital executives.

Anyone who has ever worked in a hospital knows that change is constant. However, a cyberattack is among the most disruptive and devastating events that can occur within a healthcare environment. A 2021 study by the federal Cybersecurity and Infrastructure Security Agency (CISA) found that hospitals hit by ransomware often experience additional stressors that can be correlated with higher patient mortality rates.

This can happen at any facility, but criminals shifted their focus to attacking small and rural hospitals in 2022. Even though a successful attack against a smaller facility may yield less patient data or a lower ransom to release data, the reality is that they are often easier to breach and invariably connected to larger facilities.

When an urban or suburban hospital is hit with a cyberattack, it may inconvenience patients, but they often have other care options nearby. That’s not the case for rural hospitals. The nearest facility may be 40+ miles away, which doesn’t make it feasible to simply divert patients. Even if patients are diverted, nearby facilities can become overwhelmed, creating a cascading crisis throughout the community.

The stakes couldn’t be higher, as evidenced by a 2019 attack on an Alabama hospital that knocked out the hospital’s IT systems for three weeks and is believed to have resulted in the nation’s first fatality attributed to ransomware. According to the lawsuit, patient monitors were offline while the plaintiff was in labor, leading to insufficient monitoring of a fetus that was born unresponsive with the umbilical cord wrapped around the baby’s neck. Although the child was resuscitated, brain damage occurred, and the infant died nine months later.  In a recent 2022 attack, a rural Washington State hospital was so overwhelmed that an ER nurse called 911 for help.

The urgency of improving the security posture of these small and rural facilities continues to escalate every year.

As the sophistication of cyberattacks continues to grow, the federal government should be stepping in to help secure these hospitals and keep patient data safe. As I testified to the Senate committee, implementing these four measures could improve the state of cybersecurity for our small and rural hospitals.

First, we must move beyond guidance and recommendations and create minimum standards for cybersecurity that all healthcare organizations must follow. These standards must be reasonable, effective, achievable, and continually evolving as cybersecurity requirements change over time.

Based on the items outlined in the Health Industry Cybersecurity Practices (HICP) document, recommendations can be grouped into five basic categories:

• Email security and protection
• Access management
• Asset management
• Network management
• Incident response

Simply put, regulators must spend less time suggesting and more time providing concrete solutions.

Second, we cannot leave our small and rural hospitals behind. We must create funding opportunities to allow all hospitals to meet the standards. Options include:

• Subsidies, which have found success among rural hospitals in other initiatives
• Grants, which may prove more difficult as smaller hospitals often don’t have grant-writing resources
• Incentives for small and rural hospitals to enhance security, a “Meaningful Security” type program modeled on Meaningful Use
• Enhancements in Medicare and Medicaid payments for eligible facilities, with hospitals showing how additional funds were used to boost cybersecurity

Third, we need better coordination of government cyber efforts for healthcare. While the guidance and services from government are appreciated, there is often a knowledge gap regarding the unique healthcare challenges that must be considered when applying cyber best practices in this sector. Due to time and budget constraints, many rural hospitals find it challenging to access or use available resources, so coordination must be streamlined to be effective.

Fourth, the federal government should establish a cyber disaster relief program, much like the assistance provided by the Federal Emergency Management Agency (FEMA). Such a program would provide this vulnerable sector with valuable resources in the event of attack, assist organizations in their recovery process, and increase the likelihood that hospitals could keep their doors open following a cyber-attack.

Overall hospital operating margins have been in negative territory for the past 12 months, according to a February hospital report from Kaufman Hall, and margins have decreased year over year for the past eight months. Operating margins are often higher for larger facilities that have outpatient clinics and more ancillary services than a smaller hospital can offer.

Adding to the challenging complexities, nearly 700 healthcare data breaches of 500 or more records occurred in 2022, according to the Office for Civil Rights. While the number of breaches is basically flat, the number of breached records topped 51 million for the first time, apart from the anomalous 2015, when just two breaches exposed 90 million records. Cyber insurance rates also continue to increase, with insurers demanding more monitoring and detection technologies that smaller facilities may not have if facilities can obtain insurance at all.

Because healthcare records are so valuable, hackers aren’t going to stop. Small and rural hospitals need help to protect their systems and patients, and these simple measures are a sensible path forward.

Readers Write: Turning Data into Action to Address Social Determinants of Health

Turning Data into Action to Address Social Determinants of Health
By George Dealy

George Dealy, MS is VP of healthcare applications of Dimensional Insight of Burlington, MA.

Social determinants of health (SDOH) are widely recognized as critical factors that influence the health outcomes of individuals and communities. SDOH refers to the various environmental, economic, and social factors that impact a person’s health, including food and housing insecurity, social isolation, and lack of reliable transportation options.

While most people are aware of SDOH, they often struggle with how to derive meaningful insights from the available data. Therefore, it is essential to explore practical ways to turn data into action to address SDOH and reduce health disparities.

Failing to address social determinants of health can have dire consequences, particularly for underserved populations. CDC data shows that certain populations, such as minority groups, are disproportionately impacted by SDOH. The health outcomes of these groups are often comparable to those in third-world countries.

One of the most striking examples of this is the maternal and infant mortality rates among certain minority populations. For instance, maternal and infant mortality rates are significantly higher among African American women than among other racial and ethnic groups in the United States. Addressing social determinants is a critical step in reducing these disparities and improving overall population health. This highlights the need to use data related to SDOH for actionable change.

One practical way to address SDOH is to use data and analytics. It’s first important to know where to find data related to SDOH. The data can be obtained from various sources, including national surveys, government agencies, and community organizations. One such resource is the US Census Bureau American Community Survey, which provides valuable data on community demographics, including economic and social characteristics at the neighborhood level. As SDOH information is very geographically specific, data at this level can help identify specific needs and target interventions more precisely. Another important resource is the US Agriculture Department’s Food Desert Map, which helps to identify communities where residents lack access to healthy food options.

Additionally, many non-profit organizations aggregate data from various sources into information resources. These resources include the County Health Rankings Project, run by the University of Wisconsin and the Robert Wood Johnson Foundation, and Community Commons, which is a collection of tools and resources for democratizing data related to advancing equitable community health and well-being. These data-centric resources can empower healthcare providers and policymakers with the necessary insights into the needs of the community and identify potential solutions.

The next step is to analyze the data to identify trends to better understand the specific challenges that a community is facing in relation to SDOH. One practical approach is to use technology to analyze and visualize the data. This can help identify trends and patterns more efficiently and communicate findings in a clear and compelling way.

Leveraging data on SDOH can help in the development of targeted interventions tailored to address the specific needs of each community, such as expanding access to healthy food options or improving transportation services. For instance, data mapping tools can identify areas with high rates of poverty and food insecurity, with this information then used to target interventions in those areas. Predictive analytics can forecast potential health risks based on social and economic factors. The use of data and analytics can be a powerful way to identify trends related to SDOH, develop appropriate interventions, and measure their effectiveness.

Improving population health requires a comprehensive understanding and focus on social determinants of health. While healthcare plays a crucial role, it is only one piece of the puzzle. Addressing SDOH requires a practical and collaborative approach that involves analyzing data, leveraging available resources, and recognizing the dire consequences of inaction. By turning insights into action, we can make meaningful progress in improving the health outcomes of communities, particularly among minority populations, and ensure that every citizen has access to the care they need and deserve.