Editor’s Note: Erdem Asma MSM, PMP is a healthcare executive with more than 15 years of healthcare technology implementations experience supporting business requirements for both HIS vendors and healthcare organizations globally.
“Awareness of a problem does not mean much, particularly when you have special interests and self-serving institutions in play.” – Nassim Nicholas Taleb
The term population health is how new medicine should be able to identify a population and predicting what their health needs may be via constant intervention to improve health in a better healthcare system environment. The idea is to produce a healthcare service which does not start and finish at the hospital door, although intertwines all aspects of community and primary care.
We utilize enterprise level platform systems within structured optimization services settings. From a non technical perspective these often presented as systems which is able to integrate data (weight maintenance, exercise regimes, etc.) from a variety of sources. These sources may be primary care, secondary care and even information gained from patient web portals that the patient can enter remotely in order to create an observational study of the same variables over long periods of time to gather holistic patient record. We then apply the findings to algorithms to the record to identify those at risk of disease and subsequently alert the physician – rather like a very sophisticated early warning score which we all are familiar with. This approach required the necessity of a role as health coach within the optimization services setting – someone who is able to keep an eye on a patient’s EHR and help to make behavioral changes when the system shows that someone is going off track and is therefore at risk of representing to hospital.
We often utilize the records of weight, diet and exercise which the health coach able to tell from the weight readings by the home scales. Once they are automatically synced to the health record we track that the patient was putting on weight. As the projected next level we apply the data to a model of chronic disease; asthma, cardiovascular disease, chronic pain, COPD, diabetes, glaucoma, multiple sclerosis, and stroke. The data will help patients to feel more empowered about the management of their condition. Neurologists or clinical nurse specialists could use this tool to manage the patient’s condition from home, to treat relapses, or even more importantly to predict them, and therefore ensure early treatment with DMT/immunosuppressant and perhaps enrollment in appropriate clinical trials.
Moving forward technology is necessary but changing behavior is really hard. We are where we are, and getting to where we must be will take time. Provider systems are designed to get the results they get under a heavy fee-for-service influence. Currently most providers are ill equipped to provide population-based care or manage risk. Health plan systems are designed to perform the tasks they do while fee-for-service has strongly powering them as well. Core administrative systems and processes are ill equipped to support population-based care. Regulatory environment evolved from a policy point of view that ‘more-is-better’ and competition ‘is essential’ to reduce costs . Anti-trust barriers to physician integration such as Stark Law is an obstacle.
Positive program outcomes within population health technology
– Acquire data from all sources (Medicaid Claims, EHR Clinical, etc.)
– Bring insight to providers similar to dashboards
– Develop and deploy predictive algorithms
– Drive strategy for resource allocation and clinical program implementation
– Collect Supporting Data
– DHA provider outreach and care coordination
– Deploy programs and services to providers and the communities
– EMR training and support
– Pre-diabetes/diabetes screenings
– Maternal health initiatives (Parents as Teachers, Parent Educators)
– Readmission prevention (Health Coaches)
– Digitization of clinical records in the delta
An Example of a Process Flow – Predictive algorithm for pre-term births:
- Mine the Data
- Number of providers capture semi-annual or annual data.
- Note the trend towards common clinical features
- Calculate weighting Apply regression model or machine learning.
- Define the high-risk factors
- Maternal age Multiple pregnancies or short intervals between Anxiety or depression Smoking Race Vaginal infections High cholesterol Asthma Chronic and gestational diabetes.
- Validate the algorithm
- Study subpopulation of pre-term births and trends.
- Ascribe score to a patient
- For subsequent pre-term births Establish threshold for action/intervention.
Creating Shared Goals & Initiatives…
What we know is the initial key in order to make a greater impact on Population Health. Therefore we needed to segment our patients into key focused groups. Also identify gaps in the care planning process rather than anecdotal case findings. Assess root cause to readmissions more timely.
As the next step identify, what we have such as:
– Pieces of internal data (demographic, clinical data, medication)
– historical data
– foundations for transitions of care
– cross continuum collaboration.
Then we have to focus on where we need to go by starting with risk stratification. Continue with acknowledging method to track care processes in order to drive improvement around interdisciplinary team processes. Recognizing the value in data analytics, flags, workflow design, automation, and dashboards will improve provisioning of data for cross continuum care partners and enhance post hospital follow up.
Finally how do we get there which must fundamentally change how care is delivered. We have to move from a primary focus on volume: Full census; busy MRI; booked OR; hips, knees, tests, procedures… We have to move to a greater emphasis on value: care that measures up to best systems on cost and quality metrics; patients who are engaged in their healthcare; providing the right care at the right time to the right patient.
At the end, population health at its core is the merger of financing and the delivery of healthcare. “Patient engagement” is a broad concept that combines patient activation with interventions designed to increase activation and promote positive patient behavior where EHRs are becoming commodity platforms. The winner will be the EHR vendor that provides the best platform for innovation – the most open and most extensible platform.
Foundations for the Next Decade
As the main goal monitor and improve quality to measure and achieve quality targets; providers need the ability to integrate EBM, document processes, monitor and report on quality markers.
Coordinate care and engage patients to effectively decrease costs and keep patients well, clinical data must be fluid, usable and have consumer accountability. Since payor decisions will be made across large populations, providers must have the ability to generate population-level reports and analytics as accurately as possible.
EMR vendors must be able to understand how each provider is delivering care and manage the payment distribution to incent the correct outcomes. Unlocking the full potential of the digital age in health care will require access to an efficient use of the population management data.
Opinions expressed by HIT Consultant Contributors are their own.