Harnessing Digital Health Data for Optimal Outcomes

Harnessing Digital Health Data for Optimal Outcomes unknown

PRESS RELEASE

Published May 26, 2023

Summary.

Digital tools are generating a growing amount of information about health. This information can be and should used to improve health care quality. This article explains the benefits and what is required to achieve them.

These days, we hear a lot of talk about “digital healthcare”. We ought to understand more about our health now that we have so much data, thanks to electronic health records, fitness apps, gadgets and home genome testing kits. It’s not enough to have a lot data. We must be aware of the data we have, know what it means and take action based on this understanding. The challenges in health care are not unique to the United States. They exist around the world.

This is a very common scenario.

June, 67 years old, presents to the emergency room with abdominal pain and bleeding in her rectal area. The tests reveal that the colon cancer is inoperable and has probably been growing for many years. She enters hospice after several unsuccessful and difficult chemotherapy courses. She dies a few weeks later.

Colon cancer can be largely cured and is often preventable, if caught early and precancerous growths are removed. If June had been screened, she might still be alive today. What happened? She had colonoscopies at 50 and 60 on time, but assumed she was safe until 70. No one noticed the radiologist’s note that a few minor irregularities meant she needed to come back at age 63. The radiologist was not responsible for making sure June took action on the findings, which were hidden in her EHR’s “Test Results”. She missed it. She missed it. The entire health care system missed this.

Too many Junes have been lost. The U.S. healthcare system is full of these small mistakes with huge consequences, which cost Americans millions of dollars and years of health life. It’s not surprising that these failures happened when clinicians were relying on paper folders, multipart forms and landline telephones to perform and track their tasks. Since computers, smartphones and the internet are available, they can (theoretically) be used to remind patients to have early colonoscopies.

But digital tools do not use themselves. We must tell them what to. In June’s situation, the correct combination of systems should have detected and analyzed the data, sent it to her and to her doctor, tracked their responses, made it easy for June to “click here” and schedule her procedure when she turned 63 and followed up with recommendations for testing and treatment. Although “alert exhaustion” is a real danger that must be avoided by clinicians and staff, they will appreciate reminders that are designed to help avoid missed or delayed diagnoses and regrets.

The most important and daunting mission in health care today is to figure out how to create systems that can use the growing amount and variety of digital data. Since the 1990s our organization, the National Committee for Quality Assurance, has used data to measure and enhance health care quality. Originally, this was to accredit healthcare plans, and in more recent years, to gauge provider performance. The NCQA’s first challenge was to collect enough data to be able to draw inferences and fill in any blanks. The overwhelming amount of information that must be sorted out to find its essence is now the biggest challenge. NCQA’s goal remains unchanged: To put data to use to improve the efficiency of resources allocated to health care.

This article will explain how to close the digital loop between information and action.

Basic Principles of Measuring the Quality of Products

Three questions are at the heart of measuring health care quality.

• Do we do the right thing to manage health and healthcare?
• Do we get the results we want?
• What needs to be changed?

There are almost no easy answers to these questions. People are not widgets. The outcome of an episode of care is dependent on many factors, including the performance of the clinicians, their attentiveness, the initial health of the patient, his motivation to improve, and other circumstances, such as income, location, transportation or food access, and availability of assistance around the home. The outcome of a particular episode of care depends on multiple factors, including the performance and attentiveness by caregivers, as well as the patient’s initial state and motivation to get better. Other factors include income, environment and access to food or transportation.

Although it is difficult to measure the quality of the care, we know that the United States’ current report card paints a mixed image. The best care available in the United States is often the best anywhere. However, it is primarily famous in health care circles for paying the most (greatest impact, it is important to understand the effectiveness and efficiency of medical providers and services.

The measurement of quality in health care is underdeveloped for several reasons. The first is that the share of revenue from most providers is still a minority. The second reason is that consumers haven’t demanded it. Instead, they rely on the advice of their doctors or family members who have treated the condition.

The primary reason that quality measurement is limited is because it relies on insurance claims to measure.

Claim Data: A Foundation that is Incomplete for Measuring Quality

Since the industry started a serious data-driven effort in the 1990s to measure quality of care, it relies heavily on the analysis of insurance claims. This is the only digital data source that is large and consistent across all providers. Although claims data can offer some insight, the data is usually not suitable for other uses.

It’s usually months old when it is ready for analysis. It’s also clinically insufficient. The claim only shows what was done, not its effect. The list of completed tasks – blood sugar tests and eye exams as well as weight and blood pressure checks – shows that the diabetic patient was treated but does not show whether her blood glucose is under control. The claims will not include vital information about the patient’s health, unless it makes the provider more cash. The provider can charge a higher rate for a diagnosis if the patient also has a comorbidity. For example, treating diabetes in a patient with a heart condition. Linking the patient’s claims may be the only method to determine that she has arthritis, reflux disease, and eczema.

Each claim is only a partial snapshot, a snapshot of a service or episode delivered at that moment. A pile of snapshots, however, is not a movie. Between the snap shots, health can improve or deteriorate. It’s too late by the time the photo is taken to change the outcome. All we can do then is to look at the results and consider how we could do better the next time.

The Era of Digital Measures

We no longer need to rely on data from claims. In 2010, the federal government began offering incentives to encourage electronic health records. The Office of the National Coordinator for Health Information Technology (ONCHTIT), which led this massive effort, continues to promote and initiate ways to leverage medical billing software’s information.

In recent years, this data has been supplemented by data from fitness trackers, smartphones, monitoring devices, patient assessments of their own health, genomics, and easily accessible data at the population level on factors that affect health such as employment status, income, environmental quality, community support and more. Advanced analytics could allow us to combine these data sources in order to develop a more accurate picture of health status, and the effectiveness of healthcare at all levels – from individuals with the same diagnoses to whole communities.

This is the supply side. CMS, the largest payer of health care in the United States, is actively advancing digital data as a way to measure quality of care. Since it is difficult to conduct “value based” contracts if reliable measurements are not available, commercial payers also seek better ways to measure value. Our organization is creating digital measures to track performance of health plans that we accredit. These plans collectively cover more than half of the U.S. populace. Every organization that has a stake in evaluating health care quality prepares for the new era.

Learn from others

The United States could learn from other developed nations that use their digital data for improving health care. Denmark, for instance, has patient records dating back to 1960s and a shared electronic health record system for the entire country. National Digital Health Strategy is a strategy that focuses on the same things as the United States: timely information, partnership with patients and prevention. Denmark is a smaller country with less than 6 million residents, making it a more manageable challenge than the United States.

The European Union has similar goals. In May, it presented a proposal to create a single market in digital health for its 450,000,000 citizens.

Other countries who are facing similar issues of access, affordability, and quality of care will find the United States’ efforts to promote digital measures of value and interest.

A To Do List for Digital Measures

There are four main imperatives that we see to get the United States to where it needs be:

Improve the timeliness of data collection while reducing costs.

It may seem like there are two goals, but digital methods achieve both. In some cases, traditional measures can be rendered irrelevant by using data (such insurance claims), which is often a year behind the actual care. When designed correctly, electronic health records (EHRs) and wearable devices will generate data that can be used to manage care more efficiently. Once data collection is no longer a separate process from providing care, we can move straight to the analysis and results.

Expands the range of data that can be used.

The new data sources that we have mentioned — EHRs and wearable health monitors — as well as patient feedback (also known as Patients-Reported Health Outcome Measures or PROMs — could be combined with information about the environment of the patient, such as air and water quality, crime, green space and transportation access, as well as the density of social services and grocery stores.

NCQA is looking at how to take into account the social circumstances of patients — such as homelessness, poverty or isolation — when assessing their quality of care. A physician might recommend that a person go for a walk every day. This is a good idea for someone who lives close to a park, but not for someone who lives in an area with high crime rates and fears leaving the house. We can develop more accurate measures by analyzing more data about more patients. We can account for differences in care requirements depending on the economic situation, the ability of patients to manage their care and the quality social support they receive.

Use mobile devices and artificial intelligence, as well as the widespread adoption of electronic medical records, to guide care and provide real-time feedback.

Electronic health records have evolved from being just a record of a patient’s medical condition and care received, to offering real-time assistance: alerts, notifications, computer-based guidelines on managing chronic diseases, and logic which (tactfully critiques) a doctor’s orders for testing and medication by comparing them with standard practice and checking any inconsistencies. An intelligent EHR could have reminded June’s doctor and herself to schedule her follow-up colonoscopy at age 63.

Our systems of measuring care quality will become more sophisticated as we incorporate intelligence more tailored to the individual needs and preferences of patients. An intelligent EHR will notice that June prefers to have her medical appointments scheduled on Tuesdays. With her consent, it would schedule the procedure the next Tuesday available.

Integrated health systems like the Intermountain Healthcare in Salt Lake City or Geisinger in Pennsylvania have developed digital tools for improving care. Both have the dual advantage of advanced IT capabilities as well as the financial incentive to focus more on improving the health of their patients than on simply delivering services. They and other organizations have used their electronic health records in order to give real-time feedback to patients and clinicians. These systems are able to provide more personalized feedback to patients by lowering the cost of gathering data and expanding the types of data they collect.

Create a digital platform for ongoing production processes that gather, analyze, and report quality measures.

Digital measures are not an overnight project, but rather a constant transformation. This foundation is built on the following:

Developing a standardization process for the various measures currently in use. The process must be rigorous to ensure that everyone agrees on what constitutes hypertension, or what ranges of test results indicate well-controlled diabetics, while also being flexible to allow for adjustments based on population or individual. At the moment, payers, regulators and professional associations all take slightly different approaches when it comes to designing measures. This variation is more difficult for providers to measure, but it almost certainly doesn’t deliver a commensurate amount of value.

Replace the paper-based description of quality measures, and the data that they require. This must be manually input into electronic health records or reporting software. It is best to replace paper-based descriptions with software-based ones that can easily be added to clinical systems.

Create software tools to facilitate collaboration when developing, testing and maintaining measures. Illnesses and treatments are not static and will require their own measures. In order to speed up the development of new measures, payers, regulators and providers must all participate.

Automation of the extraction from medical practice management software instead of using human data abstraction (still a commonly used practice). It will also improve the accuracy of clinical data. Fast Health Interoperability Resources is already a good tool to do this. It’s a standard API that allows information to be exchanged between systems. CMS will begin requiring providers to use FHIR enabled systems next year.

Automating auditing and cleaning of data. A large amount of data, but not all of it, in EHRs, clinical systems and other electronic health records is entered manually, which can lead to mistakes, omissions and inconsistent entry methods. Digital measures are worthless without excellent data.

Every stakeholder in the health care system has a role to play, including creating infrastructure for digital information.

• The community of quality measurement needs to expand and intensify its efforts in order to identify which new data elements will be most useful for identifying the best practices and explaining differences in outcomes.
• Hospitals and insurance companies both have legacy computer systems which struggle to support the exchange of data with other systems. To meet the needs of digital measurements, they need a combination of upgrades, standards or workarounds.
• The primary payment for physicians and hospitals is based on volume of care, not quality. This reduces the motivation to change their approach to healthcare delivery. Payment models that are based on value and effectiveness must be adopted by both payers and providers.
• Employers and governments pay the majority of the cost of health care. They have an important role to play by using their influence (e.g. contracts, their ability to move provider and health plan businesses elsewhere) to insist that health plans, providers and the community of quality measurement accelerate the adoption and development of digital quality measurements. Employers and governments can also use their talent to help the industry better understand how the measures will be used to enhance the health-care benefits they offer. Their staff should also participate in forums to define health data standards, as well as the appropriate use of data.
• Patients need these insights to be readily available in a form they can easily interpret and evaluate when making decisions about their health or health care.

Digital Measures and Their Impact

What would it be like to be able measure and manage our healthcare quality using this massive amount of data?

The providers could improve their performance by assessing it more accurately. Patients who are due for screenings would be caught, chronic illnesses that land patients in the hospital regularly if not managed could be managed and some chronic illnesses might even be prevented with strategic attention and education.

Patients and their families could make better decisions. The same digital methods used to suggest restaurants or places for oil changes could be used to help patients find the best possible care.

Insurance companies and employers can refine their health benefits to better meet the needs of employees and members. They could pay for services that keep people healthier and find the best providers. They could do this in real-time, or very close to that, rather than relying solely on last year’s data.

Health care can become a data-driven powerhouse, similar to retailing and financial services. But it will be used for saving lives and keeping people healthy.