The healthcare system in the US is complex, incentive driven, and highly fragmented. Poor coordination in a fragmented system has resulted in serious problems, such as inconsistency and inefficiencies that affect key stakeholders. It’s no surprise that a recent study by Johns Hopkins Medicine suggests that medical errors should be ranked as the third-leading cause of death in the US1, 2.
As with any complex system that has evolved for decades, the US healthcare system too has a complex and rigid structure and structural inconsistencies. Even though standards and canonical notations are widely used there aren't many unification efforts. Furthermore, information technology adoption has been slow, inconsistent and siloed, creating many interoperability issues3.
Working with customers in the healthcare space, we’ve seen some common complexities that lead to inefficiencies, inconsistencies, and medical errors in the US healthcare system.
Let’s first understand the challenges from a patient’s point of view and then discuss the key federal programs that emphasize best practices for Healthcare Information Technology (HIT). We will then look at implementation complexities, reference architecture models that can be deployed during patient care transition, drug prescription sharing, medical tests ordering and results sharing. Finally, we’ll discuss the infinite possibilities of an efficient healthcare system and the benefits its stakeholders can reap.
2. The US Healthcare Landscape - Challenges and Opportunities
If you are a US resident, the healthcare system doesn’t offer you a pleasant experience due to the hassle associated with the many layers and players you have to deal with. In a typical scenario, you have to first schedule your appointment, provide your medical history each time, schedule prescribed medical tests (and wait for the results to be sent to the physician), then pick up the drugs from the pharmacy (and wait for the physician to send the prescription), and finally deal with insurance claims and payments. This certainly is painful, but as a stakeholder you want to believe that the system is well connected and streamlined and you get the best care possible. However, the truth is far from what you will believe. The same study1, 2 that talks about medical errors elaborates the types of errors like unit mismatches, type and dosage conversion errors, misinformation during patient care transfer, wrong medical tests being performed and inconsistencies in test data while transferring are deemed the leading causes.
It is often argued that the US healthcare system incentivises the wrong actions, the physicians are paid more for doing more regardless of the end result, which is the betterment of the patient. Patient progression is not effectively checked nor followed up. There is also overemphasis on high tech care (often very expensive), but not effective care. With such incentives the US healthcare system processes close to ten billion laboratory tests a year and a hundred million CT and MRI scans, and so on4.
The US healthcare system is also provider centric and is, therefore, fragmented and uncoordinated. There is no single source of truth about a patient who is the care receiver. This siloed system is sorely mismatched to the nation’s overriding health challenge, namely, providing coordinated, integrated, continuous care to more than 125 million Americans who suffer from chronic disease5.
To address the challenges, the system too needs a ‘superdrug’ in the form of a “patient centric” structure, where the primary care is coordinated around the patient with emphasis on what the patient needs. For such a structure to function efficiently, HIT has to be the key ingredient as it connects providers and services that links the patient, thus creating a holistic view of the patient’s health and medical history.
Figure 1: Patient centered medical home
3. Institute of Medicine (IOM) Vision and Federal Programs
The IOM has further validated the poor quality of healthcare that’s linked to structural issues like the lack of an integrated system that connects all stakeholders. They provide evidence with data where patients who visited three to four doctors within a two-year period are vulnerable as the patient care is not effectively transitioned6. The IOM’s vision is to improve care quality by effectively using IT to connect providers and thereby making the system more patient centric7, 8.
Figure 2: IOM vision
To implement the IOM’s vision, the office of the national coordinator (ONC) put forward several programs that comprised of HIT directives, such as electronic health records (EHR), health information exchange, standards and interoperability, and research and data science among others.
Through incentives, the ONC encourages providers to comply with these directives and use them meaningfully (i.e. meaningful use directive).
4. Putting the Key Programs to Practice
In essence, these programs promote best practices in IT that align with healthcare provisions. EHR promotes the effective use of standard messaging formats and data dictionaries. Interoperability promotes the use of standards across the board, standards of data formats, and standards of data transfer protocols. Information exchange enables relaying data and making the data available for consumption. Last, but not least, research relates to normalizing, summarizing, and predicting with the collected data.
The ONC thoughtfully outlines the key KPIs to measure meaningful use of IT in its certification process. A few important KPIs relevant to the topic are as follows:
- Support of EHR
- Improve interoperability
- Ensure privacy and security capabilities
- Facilitate data access and exchange
- Improve reliability and transparency
4.1 Electronic Health Records (EHR)
Adoption of EHR has a major impact on quality improvements in healthcare, namely electronic prescriptions, drug to drug and drug to allergy interactions, medication reconciliation, patient reminders and automated measurement calculations. It improves coordination by cataloging electronic copies of health information, timely access, clinical summaries, and electronic patient care transition. It improves public health by enriching the national databases (public registries like cancer, reportable disease, etc.) and providing electronic surveillance for epidemics and bioterrorism.
Interoperability is a key component of effective EHR. Regardless of being digital, health records were written down with standards, such as ICD since the 1900s. With digitization, data standards, such as ICD, CPT, and SNOMED became a part of EHR and interoperability, thus forming a critical aspect of HIT.
Figure 3: Medical data, document and messaging standards
Apart from the commonly used standards, data and protocol conversions are trivial tasks in a well-formulated HIT system. Interoperability also means data (SNOMED like) are converted to document formats, then to a message (HL7/FHIR like) sent via a certain protocol compliant with a process (IHE like). The HIT tools must understand these steps and each building block along the process. One of the main reasons why interoperability is critical and integral to HIT is that HIT itself continues to evolve with new standards and protocols.
4.3 Ensure privacy and security
When we talk about EHR, interoperability, and data exchange in a connected HIT platform, focus on security and privacy cannot be overlooked. As patient health data is protected under HIPPA privacy and security rules9, the HIT has to ensure proper enforcement while facilitating meaningful use. This include data normalization and data de-identification for research and data science purposes. The HIT implementation has to concentrate on authentication and rule-based entitlement to provide access to the right person or to the right system.
4.4 Facilitate data access and exchange
Data access and exchange touch upon many aspects of the HIT platform. Modern protocols like FHIR greatly facilitate data access and exchange needs in a readable and an efficient machine processable manner. Document exchange also improves patient care transfer to a great extent. API-based data provisioning for system to system integrations has greatly improved provider-based institution integrations. Each of these exchanges has to enforce aforesaid privacy and security rules, hence in essence, data access and exchange means a well-designed API based ecosystem with characteristics like governance, self service, throttling, security, entitlement, and monitoring.
4.5 Improve reliability and transparency
Patient data often means expensive test results, care history on chronic illnesses or follow up data post a surgery. This data is highly regarded and has to be handled reliably in a transparent manner. The HIT platform should be designed with backup systems and replay scenarios during outages. It should implement messaging patterns, such as dead letter channels, to improve communication. It should also maintain logs, audit trails, and employ message tracing to improve transparency in HIT platform transactions.
5. Reference Architecture and Implementations
The outlined KPIs by the ONC provide a solid background for an HIT platform reference architecture. Digital infrastructure for learning healthcare system8 publication also outlines the use of middleware and specific technology like API management tools and identity and access management systems to create an effective HIT platform.
Figure 4: Connected healthcare reference architecture14
Well regarded reference models in service oriented architecture (SOA) use patterns like an enterprise service bus (ESB), service registries and data dictionaries to realize the HIT platform requirements15. There are proven platforms like at Spectrum Health10 that adopt these patterns to build an effective patient-centric healthcare platform.
Companies like ZeOmega are positive opportunists who have built a platform that can be a reseller to any healthcare organization looking to improve their care quality11. According to ZeOmega, they use EHR effectively by creating API-based ecosystems and by enabling secure, reliable, and transparent EHR exchange12.
Following the ONC directive, state agencies like South Carolina Department of Health (SCDH) have taken initiatives to reduce medical errors by adopting HIT through programs like South Carolina MMIS13.
As with any large scale multi-stakeholder platform, the US healthcare platform too is an evolving system. The reason it’s called the “learning” healthcare system is that it improves iteratively with learned experiences.
The opportunities that can be offered with the use of informatics in the US healthcare system vastly outweigh technological and implementation challenges. The key, however, is adopting the right technology that meet the specific requirements. That said, HIT cannot provide a silver bullet for all pitfalls of the US healthcare system;however, it can provide a solid plan to mitigate prevalent medical errors that often put patients at great risk. IOM’s directive and ONC’s incentives based initiatives are highly commendable. It shows that the healthcare governance body has correctly identified the current limitations and drawbacks of the system. The emphasis on HIT and the meaningful use of EHR add much needed reliability to critical patient data records. Emphasis on the patient centered medical home is a pivotal change to the current provider centric system. Moreover, it’s encouraging that some organizations are leading this transformation with HIT, connecting providers and services around the patient and thus creating an efficient care network.