By Chris Plance and Kunal Patrawala, healthcare experts, PA Consulting.
Artificial intelligence tools, such as natural language processing (NLP), can be applied to unstructured data to produce real-time and nuanced insights. NLP can be applied to many types of unstructured data in a provider setting, however, the ideal approach is to apply these tools to a non-clinical source of unstructured data first.
Non-clinical unstructured data provides organizations with a perfect platform to develop their skills. Two well-known sources of such non-clinical unstructured data are consumer assessment of healthcare providers and systems (CAHPS) and hospital consumer assessment of healthcare providers and systems (HCAHPS) surveys. Applying NLP to CAHPS and HCAHPS surveys can help providers improve their top-box scores, build confidence and capabilities in their data skills, increase revenue, and allow organizations to take the crucial first steps towards the journey of unlocking 100% of their data.
CAHPS and HCAHPS Surveys play an important role in today’s Healthcare environment and are a substantial source of structured and unstructured data
Similarly, private payers are also tying these metrics to reimbursement. The advent of healthcare consumerism has also led to patients extensively using tools such as CMS’s Star Ratings to make informed decisions. As a result, CAHPS and HCAHPS surveys have downstream effects on reimbursement, star ratings, and brand loyalty which makes improving top-box CAHPS and HCAHPS scores vital to a provider’s financial and operational health.
CAHPS and HCAHPS contain both structured (numerical ratings) and unstructured data (patient comments). These patient comments are accessible, and approximately 50% of the patients who take HCAHPS surveys leave comments. Unlocking this unstructured data can provide more comprehensive and specific information that cannot be gauged from only structured numerical ratings.
By Kim Huynh and Esther Ketelaars, health and life sciences experts, PA Consulting.
Studies unanimously show the negative effects medication non-adherence has on clinical outcomes and healthcare costs. Adherence is critical to ensure that medications work properly and important for pharmaceutical companies because it helps keep their drugs covered under health plans. Healthcare insurers want to ensure their covered drugs are treating their members effectively and are seeking to prevent more costly health care.
So, while traditionally medical adherence solutions have been paid for and promoted by pharma and payors, a new player has a chance to enter the scene and challenge the status quo. Now is the time for providers to proactively address medication adherence as they start to bear more financial risk through value-based care models.
How can providers play a larger role?
Provider organizations who focus on medication nonadherence have a great opportunity to improve patient outcomes and support their value-based care models. According to the Center for Disease Control medication nonadherence results in 10% of hospitalizations and 125,000 preventable deaths in the U.S. each year. Likewise, low medication adherence leads to treatment failures between 30% and 50% of the time.1 These negative impacts become even more relevant to providers as patient outcomes continue to be more closely tied to reimbursement and payment incentives.
Pharmaceutical companies, payors, some providers, other consumer-focused companies in the healthcare value chain, and even governments have tried to address nonadherence. With the recent growth of digital health solutions especially, many have focused on leveraging novel technologies. However, there are a myriad of reasons for patients not compliant with their medication and adopting a single tool or technology has rarely been effective in reducing nonadherence.
A one-size-fits-all approach will often only address a single issue for a limited number of patients. With their direct access to patients, providers have a better chance of addressing the complex mix of reasons for nonadherence and design medication adherence programs customized to each patient. Personalized intervention plans invite a direct solution to each patient’s reasons for nonadherence with the appropriate tools that address the underlying cause for that individual.
Developing such a program requires factoring in root causes and reasons for nonadherence, using predictive analytics to identify high-risk patients, and gathering a diverse set of interventions to address those root causes. The position and shared decision-making power between providers and consumers indicate that providers can address medication adherence for more patients, improve patient outcomes, and reinforce value-based care.
Designing a patient-centered medication adherence program
Many of the tried and tested programs have been designed based on an assumption of the underlying reason for nonadherence. However, for these programs to be truly effective provider organizations need to understand and diagnose the patient’s reasons for nonadherence and tailor their intervention with the right tools.
Understand the reasons for medication nonadherence
Exploring the factors contributing to nonadherence allows organizations to understand the inherent complexity of nonadherence. Most interventions fail to produce the desired results because they don’t consider the many contributing factors. The reasons for nonadherence go beyond simple forgetfulness. Only 30% of patients cite forgetfulness as the cause of their nonadherence.
By Scott E. Rupp, publisher, Electronic Health Reporter.
The healthcare technology world is ablaze, on FHIR. New proposed standards for interoperability are being established to allow health systems the ability to share information and facilitate patient access to data. Specifically, in large part through a structure known as FHIR.
This “FHIR” the market is speaking of is Fast Healthcare Interoperability Resources, an interoperability standard for electronic exchange of healthcare information. FHIR was developed by Health Level Seven International (HL7), a not-for-profit that develops and provides frameworks and standards for the sharing, integration and retrieval of clinical health data and other electronic health information.
FHIR emerged in 2014 as a draft standard for trial use to enable health IT developers to more quickly and easily build applications for EHRs and to exchange and retrieve data faster from applications. FHIR soon received support from EHR vendors like Epic, Cerner and AthenaHealth. Shortly thereafter, the Argonaut Project emerged to move FHIR forward, and in February 2017, FHIR became a full data exchange standard.
FHIR is interoperability
FHIR is built on the concept of interoperability and modular components that can be assembled into working systems to try to resolve clinical, administrative and infrastructural problems in healthcare.
FHIR provides software development resources and tools for administrative concepts, such as patients, providers, organizations and devices, as well as a variety of clinical concepts including problems, medications, diagnostics, care plans and financial issues, among others. FHIR is designed specifically for the web and provides resources and foundations based on XML, JSON, HTTP, Atom and OAuth structures.
FHIR can be used in mobile phone applications, cloud communications, EHR-based data sharing and among institutional healthcare providers.
According to HL7, FHIR aims to simplify implementation without sacrificing information integrity. FHIR “leverages existing logical and theoretical models to provide a consistent, easy to implement and rigorous mechanism for exchanging data between healthcare applications. FHIR has built-in mechanisms for traceability to the HL7 RIM and other important content models. This ensures alignment to HL7’s previously defined patterns and best practices without requiring the implementer to have intimate knowledge of the RIM or any HL7 v3 derivations.”
Health sector buy-in
The healthcare sector has clearly bought into FHIR, primarily because of interoperability challenges.
“Sharing data between different health systems has required significant investment of IT resources on one-off projects,” said Nilesh Chandra, healthcare expert at PA Consulting. “As the needs for data sharing have increased, hospital IT departments have been swamped with demand for all of this custom integration.
“FHIR and similar standards are an attempt at standardizing data integration, to make it easier to connect EHR systems and easily extract or upload data into them, based on reusable IT components,” added Chandra. “That said, FHIR is an important step in the right direction, but is not the panacea for all health IT integration issues.”
FHIR uses a set of commonly used medical ideas termed as “resources.” The resources are used across many different types of companies and organizations, but can all mean the same thing. An example would be blood pressure readings, or an MRI scan. Those resources are held in EHRs, smartphones, health information exchange databases and so on. FHIR also allows for the mining of those elements since they are tagged in a similar way in the FHIR standard.
“The complex part is done by individual systems that don’t have the same operating system,” said Jason Reed, PharmD blog founder. “Because they can pull that tag then they pull it and exchange it with other entities. They only show the tag and not the other code or structures they had to use to get to that tag.”
While consolidated clinical document architecture allows a group of healthcare items to be sent together, this is essentially like sending an electronic PDF, Reed said. Other systems that have different operating systems can’t break that down unless they use the same operating system.
All of this is a culmination of the fact that digital health data can improve outcomes and lower costs, but the reality has been something less than ideal. For example, during the economic stimulus in 2009, systems were designed before modern web standards for storing and exchanging data were ubiquitous. The industry was caught in the middle of a technical revolution and spent its cash before the best new practices were available, said Nick Hatt, senior developer at Redox.
By Meghan Marx and Jenna Phillips, healthcare experts, PA Consulting.
Now that the 2019 health insurance Open Enrollment period has concluded, healthcare industry stakeholders are watching closely for any shifts in consumer behavior regarding enrollment and the overall uninsured rate. After a record drop in the uninsured rate from 2008 to 2016 in response to the Affordable Care Act (ACA), the uninsured rate has begun to increase again, by 1.3 percentage points from 2016 to 2017, according to data from Gallup and Sharecare reported on the Health Affairs blog. Additionally, at the December 15 close of the 2019 open enrollment period, the Centers for Medicare and Medicaid Services announced that 300,000 fewer individuals enrolled for 2019 than for the previous year. Under the policies of the current administration, the trend of decreased enrollment in exchange market plans, coupled with an increase in the U.S. uninsured rate, is expected to continue.
The uninsured rate has fluctuated over time, but factors like the Congressional repeal of the individual health insurance mandate and reduced funding for health insurance enrollment counselors to support consumers to enroll in health insurance mean that the uninsured rate is expected to climb further in 2019. The increases in the rate of uninsured may seem small, but a 1.3 percent increase in the number of uninsured Americans represents 3.2 million more adults who were uninsured in 2017, compared to 2016. While the upward trend in the uninsured rate is consistent across population segments, a few specific demographics are at especially high risk of becoming uninsured, including young adults, low-middle income earners, and Hispanics.
Here’s what we know:
Young people are less likely to purchase coverage than older adults, who tend to have higher healthcare utilization. The 18 to 25 age bracket was the least likely demographic group to have health insurance in 2017, while those aged 65 and older were the most likely group to be covered. Notable is the 2.9 percent rate difference in coverage of individuals who are 25 to 26 years old, the age when parent’s health plans end coverage, an option guaranteed under the ACA.
Lower rates of employment and education are also associated with lower rates of health insurance enrollment. Low wages for unskilled labor are typically paid on an hourly basis, rather than as an annual salary, and employees in this situation may struggle to secure work with healthcare benefits. As a result, there is a lower rate of access to employer-sponsored coverage. While the poorest individuals and families qualify for government-sponsored coverage, there is a sizeable segment of the population that cannot afford to purchase commercial insurance but is disqualified from Medicaid based on income or personal assets.
In 2017, Hispanic Americans had the lowest coverage rate of any U.S. demographic, at 83.9 percent, and the lowest rate of private insurance coverage, at just 53.5 percent. This gap in the insured rate was especially apparent in states where Medicaid has not been expanded to cover the undocumented population. Language and cultural barriers and economic disempowerment might also have exacerbated the disparity of coverage between Hispanics and other racial groups.
The Department of Health and Human Services (HHS) filed its annual year-end report to Congress at the start of 2019. The 22-page report summarized nationwide trends in health information exchange in 2018, including the adoption of EHRs and other technologies that support electronic access to patient information. The most interesting takeaway has to do with the ever-elusive healthcare interoperability.
According to the report, HHS said it heard from stakeholders about several barriers to interoperable access to health information remain, including technical, financial, trust and business practice barriers. “These barriers impede the movement of health information to where it is needed across the care continuum,” the report said. “In addition, burden arising from quality reporting, documentation, administrative, and billing requirements that prescribe how health IT systems are designed also hamper the innovative usability of health IT.”
To better understand these barriers, HHS said it conducted multiple outreach efforts to engage the clinical community and health IT stakeholders to better understand these barriers. Based on these takeaways, HHS said it plans to support, through its policies, and that the health IT community as a whole can take to accelerate progress: Focus on improving interoperability and upgrading technical capabilities of health IT, so patients can securely access, aggregate, and move their health information using their smartphones (or other devices) and healthcare providers can easily send, receive, and analyze patient data; increase transparency in data sharing practices and strengthen technical capabilities of health IT so payers can access population-level clinical data to promote economic transparency and operational efficiency to lower the cost of care and administrative costs; and prioritize improving health IT and reducing documentation burden, time inefficiencies, and hassle for health care providers, so they can focus on their patients rather than their computers.
Additionally, HHS said it plans to leverage the 21st Century Cures Act to enhance innovation and promote access and use of electronic health information. The Cures Act includes provisions that can: promote the development and use of upgraded health IT capabilities; establish transparent expectations for data sharing, including through open application programming interfaces (APIs); and improve the health IT end user experience, including by reducing administrative burden.
“Patients, healthcare providers, and payers with appropriate access to health information can use modern computing solutions (e.g., machine learning and artificial intelligence) to benefit from the data,” HHS said in its report. “Improved interoperability can strengthen market competition, result in greater quality, safety and value for patients, payers, and the healthcare system generally, and enable patients, healthcare providers, and payers to experience the promised benefits of health IT.”
Interoperability barriers include:
Technical barriers: These limit interoperability through—for example—a lack of standards development, data quality, and patient and health care provider data matching. Addressing these technical barriers by coordinating to establish the technological foundation for standardizing electronic health information and by promoting exchange of that information can considerably remove these barriers.
Financial barriers: These relate to the costs of developing, implementing, and optimizing health IT to meet frequently changing requirements of health care programs. The cost to adjust health IT to meet these requirements can impact innovation and the timeliness of technical upgrades. Specific barriers include the lack of sufficient incentives for sharing information between health care providers, the need for enhanced business models for secondary uses of data, and the current business models for health systems or health care providers that do not adequately focus on improving data quality.
Trust barriers: Legal and business incentives to keep data from moving present challenges. Health information networks and their participants often treat individuals’ electronic health information as an asset that can be restricted to obtain or maintain competitive advantage.
Elsewhere, the Center for Medical Interoperability, located in Nashville, Tenn., is an organization that is working to promote plug-and-play interoperability. The center’s members include LifePoint Hospitals, Northwestern Memorial Healthcare, Hospital Corporation of America, Cedars-Sinai Health System, Hennepin Healthcare System, Ascension Health, Community Health Systems, Scripps Health, and UNC Health Care System.
Its mission is “to achieve plug-and-play interoperability by unifying healthcare organizations to compel change, building a lab to solve shared technical challenges, and pioneering innovative research and development.” The center stressed that the “lack of plug-and-play interoperability can compromise patient safety, impact care quality and outcomes, contribute to clinician fatigue and waste billions of dollars a year.”
More interoperability barriers identified
In a separate study, “Variation in Interoperability Among U.S. Non-federal Acute Care Hospitals in 2017,” showed additional difficulty integrating information into the EHR was the most common reason reported by hospitals for not using health information received electronically from sources outside their health system. Lack of timely information, unusable formats and difficulty finding specific, relevant information also made the list, according to the 2017 American Hospital Association (AHA) Annual Survey, Information Technology Supplement.
Among the explanations health systems provided for rarely or never using patient health information received electronically from providers or sources outside their health system:
Difficult to integrate information in EHR: 55 percent (percentage of hospitals citing this reason)
Information not always available when needed (e.g. timely): 47 percent
Information not presented in a useful format: 31 percent
Information that is specific and relevant is hard to find: 20 percent
Information available and integrated into the EHR but not part of clinicians’ workflow: 16 percent
Hospitals, when asked to explain their primary inability to send information though an electronic exchange, pointed to: Difficulty locating providers’ addresses. The combined reasons, ranked in order regardless of hospital classification (small, rural, CAH or national) include:
Difficult to find providers’ addresses
Exchange partners’ EHR system lacks capability to receive data
Exchange partners we would like to send data to do not have an EHR or other electronic system to receive data
Many recipients of care summaries report that the information is not useful
Cumbersome workflow to send the information from our EHR system
The complexity of state and federal privacy and security regulations makes it difficult for us to determine whether it is permissible to electronically exchange patient health information
Lack the technical capability to electronically send patient health information to outside providers or other sources
The report also details other barriers related to exchanging patient health information, citing the 2017 AHA survey:
Greater challenges exchanging data across different vendor platforms
Paying additional costs to exchange with organizations outside our system
[Need to] develop customized interfaces in order to electronically exchange health information
“Policies aimed at addressing these barriers will be particularly important for improving interoperable exchange in health care,” the report concluded. “The 2015 Edition of the health IT certification criteria includes updated technical requirements that allow for innovation to occur around application programming interfaces (APIs) and interoperability-focused standards such that data are accessible and can be more easily exchanged. The 21st Century Cures Act of 2016 further builds upon this work to improve data sharing by calling for the development of open APIs and a Trusted Exchange Framework and Common Agreement. These efforts, along with many others, should further improvements in interoperability.”
What healthcare leaders are saying about interoperability
While HHS said it conducted outreach efforts to engage health IT stakeholders to better understand these barriers, we did too. To further understand what’s currently going on with healthcare interoperability, read the following perspectives from some of the industry’s leaders. If there’s something more that you think must be done to improve healthcare interoperability, let us know:
Healthcare technology continues to proliferate the sector, the developments almost too many to track. The sector abounds with innovation and push forward in the name of better – even the minutest – advancements of care and better care outcomes. The coming year will be no different. As we enter the final year of the 21st century’s second decade, we’ve witnessed a tremendous amount of evolution in just 19 years. What role will our healthcare technology play in the healthcare industry in the next year?
A lot. And not just for a few, but members of many, many areas, even those peripherally involved with the boundaries of care. We must understand where current innovation is, but also the challenges these migrations attempt to solve. Being aware of the trends ahead can give us all a better grasp of how care delivery is changing and we can better understand how new areas can resolve real industry problems.
To help us navigate the year ahead for healthcare and its technology, the following are some of the trends that it leaders, observers, insiders, consultants and investors think are important or need to be taken notice of in 2019.
Everybody knows that the US healthcare system is in trouble. Issues ranging from cost, to quality and access of care are rampant and only getting worse. On a macro level The Affordable Care Act (ACA) has solved some of the previous access issues, but has added tremendous cost within the system, and at the same time it has not solved the quality issues that exists.
Research suggests that the cost situation is becoming increasingly worse, which is causing firms to scramble for viability. Waves of cost cutting efforts have led payers and providers to capture some, but not nearly enough of the costs necessary for long-term survival.
There are two main cost challenges that both healthcare payers and providers share:
Wildly inefficient operating models and processes. The Harvard School of Public Health projects that of the $2.8 trillion the US spends on healthcare each year, 30 percent or $840 billion may be wasted. For organizations that function on small operating margins, this alone represents the boundary between success and failure.
Large stranded infrastructure and costs combined with declining revenues – The ratio of hospital expense vs. revenue has increased from just under 15 percent in 2011 to nearly 30 percent in 2014 with 25 percent of hospitals reporting an operating loss. For nearly 49 million enrollees in Medicare, hospitals receive only 88 cents for every dollar with lower reimbursement rates predicted in the future.
These pressures have led organizations to make hasty decisions about how to fundamentally solve the problem. Merger and acquisition activity among both payers and providers is at an all-time high, and the ACA appears to have been the catalyst for this M&A activity. Since its enactment, hospitals started merging with competitors at unprecedented rates. In 2009, pre-ACA, there were 52 announced transactions involving 80 hospitals. That number more than doubled by 2012, with 107 announced transactions involving 244 hospitals. The M&A frenzy among healthcare payers has also increased with Anthem’s announcement to acquire Cigna, and Aetna’s acquisition of Humana. Both of these were announced last year and are two of the largest payer M&A deals in history.
At HIMSS this year, multiple speakers laid out visions for a future where parents could consult with a pediatrician via a telemedicine encounter during the middle of the night, take their children to receive immunization shots at a retail clinic, and have all of this information aggregated in their primary care provider’s record so that providing an up to date immunization record at the start of the next school year is as simple as logging into the PCP’s patient portal and printing out the immunization record. In short, multiple speakers presented visions of a truly interoperable future where patient information is exchanged seamlessly between providers, healthcare applications on smartphones, and insurers.
While initiatives such as the CommonWell Health Alliance, Epic’s Care Everywhere, and regional health information exchanges attempt to address the interoperability challenge, these fall short of fully supporting the future vision described above. Today’s solutions do not address smartphone applications and still require manual intervention to ensure that suggested record matches truly belong to the same patient before the records are linked. This process is costly but manageable in an environment where a low volume of patient records are matched between large provider organizations. In a future world where patient data is available from a multitude of websites, smartphone applications and traditional healthcare organizations, it would be cost prohibitive to manually review and verify all potential record matches.
Of course, one solution to this dilemma would be to improve patient matching algorithms and no longer require manual review of records before they are linked. However, for this to be possible, a standard set of data attributes would need to be captured by any application that would use or generate patient data. In a 2014 industry report to the Office of the National Coordinator for Health Information Technology, first name, last name, middle name, suffix, date of birth, current address, historical address, current phone number, historical phone number, and gender were identified as data attributes that should be standardized. Many of the suggestions in this report were incorporated into the Shared Nationwide Interoperability Roadmap that the ONC released in January 2015.