A key challenge to making full use of your Patient Satisfaction Survey Data is learning how to integrate it with your EHR data. Having an understanding of a patient’s associated encounters can provide valuable context on the results you receive on a Patient Satisfaction Survey. The challenge oftentimes is having the right data to link these together.

Here are 5 strategies to enable integration and fully leverage EHR and Patient Satisfaction Survey Data.

1. Use MDM to match patient information
   • Use the patient submitted information to match the patient record in the EHR
2. Ensure pass-through of encounter information to your survey provider
   • If the encounter is captured as part of the Patient Survey, then ensure that you receive this your survey provider for integration with clinical records
3. Create a complete patient record from the data warehouse
   • This aggregates the encounter(s) relevant information for the patient to provide context to the survey responses
4. Group surveys by service location and practitioner
   • Using the integrated EHR data, then group survey results by location and practitioner to determine trends
5. Use integrated data to drive change
   • Identify leading indicators to unfavorable patient satisfaction scores and then use these to drive change in the organization

Using patient survey results in conjunction with EHR records can remove operational blind spots that can exist when using Patient Survey Data solely. When fully integrated, you can move on to identify trends. This then should be used to promote change and improve patient experience and well-being.

Last September I started a blog series on data integration for Healthcare. The problem, it is impressively difficult to integrate data from one or more EMR systems into a cohesive analytical database. The fact that I’ve been doing that job for the last 2 months is a testament to that statement since it kept me so occupied I’v been unable to add to this series until now.

Let me recap part 1 “Data Integration: Taming the Beast of Healthcare”.

I told you the process of extracting data from your EMR system with its associated ancillary systems and getting it to the point of analysis and reporting can be a terrifyingly complex and time-consuming task. Not to mention extremely expensive.

You need to extract the data, consolidate, clean and verify its accuracy then load it into a well-crafted data warehouse so you can generate required and usable reports.

I hinted that it could be done in less time and reduced cost if you found someone with source system expertise and a proven repeatable process and tool set to do the job for you.

Well then… let me introduce you to the Perficient Analytics Gateway for Healthcare.

Over the past few years the Perficient Healthcare BI team has been working to build a repeatable and cost-effective methodology and tool set that lets us come to your organization, pull data from almost any source system(s) and load it into an advanced and highly scalable data warehouse. We have the experience and expertise to install, customize and implement the data model (you have to buy that) and our Gateway and Accelerator product.

We work with you to get you started then train and hand the reins to your IT staff to continue to build and refine your warehouse. But I’m getting ahead of myself.  Let me explain how this works.

We designed it to work in four parts. They are:

1. A stable Industry standard Unified Healthcare Data Model in two views, an Atomic relational database model and a Dimensional star schema database model.
2. The Gateway product that extracts the data from your source system(s) and loads the data into a pre-stage database to compare with previous extracted data to identify the deltas.
3. Programs to pull the data from the Gateway pre-stage and load it to the Atomic
4. Programs to pull the data from the Atomic and load it to the Dimensional

The first part was a stable comprehensive data model that provided a complete picture of Provider (Clinical), Payer (Claims) and Accounting (Financial) data. This data model needed to be designed in such a way as to be able to keep up with the Healthcare industry’s constantly expanding data needs and structured consistently across all subject areas.

It had to have a staging database (atomic relational design). This is where you load the extracted data to cleanse and standardize formats of matching data from multiple sources. It also had to have a synchronized warehouse database (star schema design) where you perform analytics and run reports so your organization can make informed decisions to improve patient care.

We found one that was built based on HL7 data and definitions structured using a *Data Vault modeling design method. It met our design requirements and it was an industry wide design being built and maintained by one of our partners.

The data vault method is unique. In contrast, a typical relational modeling design has tables with primary key columns and all associated content. A primary key provides unique data for storage and retrieval of a record. With each retrieval from the database a row in the table is returned with all columns both key and content.

In a data vault design the primary key columns are managed in a separate table without associated content. This is referred to as a Hub. The hub contains unique business keys with additional audit or descriptive fields related to Hub columns. The Hub is stable and generic enough so that it is highly unlikely it will ever change since no data content is stored there.

The content columns are in a separate table called a Satellite which is relationally attached to its Hub. If any changes or expansion is required for content, the “navigational structure” of the database is unaffected. Satellite content changes require only minimal programming effort.

Another important component of the Data Vault design is a Relationship table called a “Link”. A Link joins Hubs together. The Link is bi-directional and always many-to-many to handle any configuration requirement. This makes it an ideal long term design as show below.

There are a few additional Data Vault Model Types but this is enough to get the general idea.

Hub and Satellite Combination.

The next cool thing is that each Hub and Satellite combination is single subject oriented or **Normalized.  Patient data, Hub A & Satellite A1, is separated from Encounter data, Hub B & Satellite B1, but they are related (linked) by the Link table (A to B) so all of the encounters for a given Patient can be viewed and vice-versa

The point I’m trying to make is this structure is ideal for navigational stability which makes it ideal for subject matter data storage and reusable programming. It also means that regardless of the source system database structures or flat file formats, we can store that extracted data in this format.

In addition, the next step in the warehouse development process (Star Schema) is also provided and synchronized with the atomic model shown above. The main warehouse tables in a star schema are Fact, Dimension and Relationship Tables.

Fact tables consist of the measurements, metrics or “facts” of a business process. Fact tables typically contain two types of columns those that contain the facts of the business process and the foreign keys of dimension table(s).

In this model Dimension tables are “Conformed” meaning they are subject oriented, timestamp based and kept in sync with the matching iterations of their atomic hub & satellite counterparts.  They are a single representation of that subject at any given point in time and used to support data across multiple Fact tables.

Using these backdrop models we built the Perficient Analytics Gateway for Healthcare. Since the staging model is stable and synced to the warehouse model we pre-built ETL programs to move the data from the Atomic to the Dimensional. Data stored in the Atomic is automatically moved into the Dimensional with little additional programming and only when customization is required.

We come to your shop, install the database, fire up our Perficient Analytics Gateway for Healthcare accelerator modules and BAM! Instead of taking 18 to 24 months to design a warehouse and build programming to load source data we’ve “accelerated” the process to get the data ready for analysis and reporting considerably faster. Saving time and money in the process.

There is one more key part to the Perficient Analytics Gateway for Healthcare though. That’s the part where we’ve developed a process to analyze your sources, map source data (any database) to the Data Vault database and process it through the “Gateway”, the crown jewel of the package.

Tune in for “Data Integration: Taming the Beast of Healthcare – Part 3” and I’ll tell you more about that part!

Till then…

*Data vault modeling is a database modeling method that is designed to provide long-term historical storage of data coming in from multiple operational systems.

**Normalized is a database modeling term that splits each subject data into its own table(s). In other words, in a Patient Encounter the Patient specific data is stored in a table(s) and the Encounter data is stored in another set then the Hubs are linked together.  This means that the Patient specific data is reusable and only needs to be stored once where many Encounters can be related to that single Patient record.

I just got back from the doctor whose office is in one of the leading hospitals in the United States. I was his first appointment of the day.

“They just updated Epic this morning, and everyone is getting into the office trying to make sense of the changes,” the doctor said. I asked if he knew that changes were coming. He said someone came around several weeks ago and told employees that a new release will be installed and that there would be someone available to answer questions from employees…at some point.

The doctor seemed a bit frustrated because the screens and process he’s been so used to have changed. At first glance, he said putting in orders for patients takes several more steps. The user experience has changed for doctors and other hospital employees. You could hear the chatter around the office, and everyone seemed annoyed.

The doctor said he just needs to sit down and figure out what changed. And, he’s right. He needs to take some time and explore the new version.

But as I was talking to him, all I could think about was organizational change management. If they had a good change management plan, it would have likely eliminated the frustration and allow employees to focus on their patients.

The four components to a good change management plan:

1. Define the change: Explain the change and why it’s important
2. Communicate the change: Tell the “what, when, how, and why” to the affected groups
3. Enable the change: Make sure employees have the training and support required
4. Sustain the change: Develop and share best practices, and adjust where necessary

Changes to EHR systems are inevitable. However, you need to plan accordingly. Time and time again, we see situations in which change management could help improve the user experience, increase adoption, and improve ROI. It’s time to take change management seriously and make sure it’s part of all your IT projects.

Let us know how our change management and healthcare teams can help you.

Healthcare providers need to create useful, not to mention required, reports and they need to perform complicated analysis. To do that effectively requires bringing disparate information from multiple operating systems together. The goal… find opportunities to improve patient care and lower operating costs. To accomplish this, operational data has to be extracted and integrated.

Integrating clinical data at this level of detail requires a (wait for it) monstrous effort. To satisfy these complex reporting and analysis requirements requires finding the needed data in many operating systems then merge it all together in a usable way. Let’s see what it takes.

To load and maintain the efficacy of clinical information in a warehouse requires expert knowledge of source systems and a deep understanding of warehouse structure and design. Lack of expertise in these areas can adversely affect the quality, accuracy and usability of a data warehouse.

Many companies have faced this challenge. Some succeed while others fail but failure means money down the drain. Bad, missing or inaccurate data may impact important decisions and even threaten proper patient care. This is not the place to skimp on expert resources.

To start, buy or build a database(s) that will satisfy the necessary reporting and analytical requirements. It should be a clean, accurate well thought out standardized design with complete, normalized and dimensional Provider and Payer data patterns. It must be subject oriented, multi-tenant and a highly extensible data structure.

Now, design an intake process that can receive data from any Electronic Medical Record (EMR) or Electronic Health Record (HER) system or other ancillary systems. The intake design would have to be generic so it could absorb data from any system.

Then, design programs to transform and load the raw data into a staging area for validation and cleansing. Following the data model pattern, design programs to move the now clean data into a normalized integrated database.

Finally, design programs to move data through the normalized model structure to a highly functional, turn-key but customizable, dimensional (star schema) data base.

By aligning the design of intake, data load, transformation and data movement processes to the model pattern, you can use this tool set over and over again. Now that I think of it, this looks like something to build and sell!

No, this would take years to build, refine and package for sale.

But what if someone has packaged these components into a healthcare reporting and analytical data warehouse integration tool set. If they built it, it would decrease build time, complexity and provide the acceleration needed to get your warehouse on-line in less time (and less cost) than it would take to do it from scratch.

What if they had a group of experts in healthcare data integration? That would be cool. Made it a team that has experience and knowledge of the top tier operational healthcare systems. People with intimate knowledge of how the packaged integration tool set operates. They would have years of experience implementing this system in multiple locations around the country.

Now that I think of it, I might just know about something just like that and I think I know just where to look!

Tune in next time and I’ll tell you all about it!

Till then…

Researchers with the U.S. Army claim that clinical data visualization is a key component in the usage and delivery of EHRs because it:

• Presents data in a pleasing, easy-to-understand manner
• Makes digesting and sharing data faster and simpler
• Reduces the time needed to interpret and understand data points
• Increases the user’s ability to see trends and patterns in patient data

Data analytics and visualization tools also allow for the smooth communication between practitioners and patients. With the assistance of communicative charts and interactive dashboards, users can absorb information more efficiently and use it more effectively.

The design and format of the information presented are crucial factors that should be taken into consideration when producing visual aids to explain the significance of the data collected.

Ultimately, analytics solutions allow healthcare practitioners to harness a better understanding of patients’ health, the available treatments, and the expected procedure outcomes. They also reduce the time spent on managing data and instead increases time spent putting the data to work for organizations and their patients.

The financial benefits of analytics solutions cannot be underestimated.

We recently published a guide that explores how data and technology can enable organizations to make informed healthcare decisions, produce better patient outcomes, and create a better patient and stakeholder experience. You can download it below.

This blog was co-authored by Tom Lennon.

In healthcare, most data is exchanged electronically between partners via EDI (Electronic Data Interchange), and “Big Data” is helping the industry become more efficient and productive. EDI originated because it provided a structured mechanism for sharing data between disparate organizations and systems.

The more common means of transferring data from source to a data warehouse is ETL, but there are times when you might want to consider using ELT (see James Serra’s “Difference between ETL and ELT” or Daniel Harris’ “ETL vs. ELT: How to Choose the Best Approach for Your Data Warehouse”).

EDI allows the healthcare industry to bring in information needed to help perform analytics. However, in the past, the issue was being able to house all the information and easily retrieve it. There are still some issues with EDI regarding data quality, but that is getting better as each business is learning the need for reliable data to perform their analytics.

EHRs (Electronic Health Records) also play a big role in the ability to perform analytics, and there is an immense amount of raw data available in EHRs, EMRs (Electronic Medical Records) and EDI. “Big Data” now provides a greater opportunity to use this information to perform critical analytics by applying business intelligence techniques.

In the past, the EDI data went to a data warehouse. Now with “Big Data,” the industry is able to house and analyze the information for visibility and quality. When linked with the adjudication system, organizations can get a more complete view of what is happening in their business and deliver real-time analytics of clinical, financial, as well as fraud and HR.

Analytics allow for the examination of patterns to determine how care can be improved while reducing the need for repetitive hospital stays and limiting excessive spending for testings etc. This allows the healthcare industry to reduce fraud, waste, and abuse. “Big Data” allows them to store and go back in their data history to analyze large unstructured datasets to detect anomalies and patterns.

Stay tuned for “How EDI Relates to Cloud Computing.”

Exposing patient health information as Application Program Interfaces (APIs) is one of the most critical components in Stage 3 of the EHR Incentive Programs and all providers will be required to comply with MU3 requirements by 2018. The APIs will ensure improved patient engagement by providing data access in application of patient’s choice instead of current patient portal channel only.

In compliance with HIPAA privacy and security rules, the implementation of APIs that expose sensitive PII/PHI data must be properly protected in terms of Authentication, Authorization, Audit, Message and Transport Level Security, Encryption etc.

Some of the key requirements for the API interface are:

1. Uniquely identify a patient and return an ID or other token that can be used by an application to subsequently execute requests for that patient’s data.
2. Allow a request for “all” the patient data, or specific “by specific data category” i.e. demographics, labs, procedure, medications etc.
3. Must include accompanying documentation that contains API syntax, function names, mandatory and optional parameters, methods and their returns, and terms of use.
4. Documentation must be available via publicly accessible hyperlink
5. Authentication, Access Control, Authorization , Auditable Events and Tamper-Resistance — Trusted connection — Auditing actions on health information

Solution using IBM APIC:

1. APIC security policies jwt-validate or validate-usernametoken can uniquely identify a patient and use of JWT/OAuth token management system through DataPower can return token to requesting client for subsequent requests.  
2. APIC is an integrated creation, runtime, management and security foundation for enterprise grade API’s, and Microservices to power modern digital applications. Fine-grained services based on data category demographics, lab results etc. can be quickly built with standards-based visual API spec creation in Swagger 2.0.
3. APIC provides customizable, self-service developer portal for publishing APIs. Through the portal, application developers can access APIs directly and can take sample code snippet from the portal to implement APIs in their applications.
4. APIC self-service developer portal can be accessed publicly for API subscription and documentation.
5. IBM APIC provides access control over API’s, API Plans and API Products. API runtime environment – API gateway (Micro gateway or Datapower) acts as Policy Enforcement Point, and maintains security and control to APIs. Datapower’s AAA policy can support variety of Authentication/Authorization formats JWT, LTPA, SAML,WS-Security, OAuth etc. Datapower should also be used to provide message level security by encrypt/sign and transport through the use of SSL/TLS.

Recruiting subjects for clinical trials has been a longstanding challenge for pharmaceutical and research organizations. According to Dr. Steven Alberts, Chair of Medical Oncology with the Mayo Clinic, only 5% of cancer patients ever enroll in a trial. On top of that, only a fraction of all trials ever finish enrolling enough patients on time and, in some cases, they never even get off the ground due to poor enrollment. More companies will begin leveraging new technology to match potential subjects to clinical trials. 

For example, IBM’s Watson can help analyze a patient’s medical and clinical attributes, compare them to various study protocols, and then determine patient/trial eligibility. Watson can even indicate the particular qualities that may have excluded the patient, giving clinicians the opportunity to help increase eligibility.

Other applications, such as TrialReach, will play a larger role in clinical trial recruitment efforts. TrialReach is a company focused on helping patients find the clinical trials they’re eligible for, based on a series of questions. Appealing directly to the public and empowering patients to take an interest in their own health research can be an effective method of improving clinical trial enrollment.

Recruiting subjects at point-of-care will likely become a focus area for academic medical centers. One scenario could look like this: A patient’s EHR record matches a specific trial’s inclusion criteria, so an alert pops up on the provider’s computer, at which point the provider discusses the trial with the patient. If the patient is interested, the provider clicks a button that sends a message to the study coordinator, who then connects with the patient, obtains consent, and completes enrollment. A similar recruitment approach can also be applied to patient portals.

To learn about other new trends that we can also expect to see in 2016, fill out the form below or click here.

Direct secure messaging (DSM) is a transmission standard promoted by the Office of the National Coordinator for Health Information Technology that meets the Meaningful Use Stage 2 requirements of electronic health records (EHRs). It works much the same way as regular email, but the message is encrypted, which prevents unintended use of the protected health information that is included within. DSM can be used to send patient information among physicians, among provider organizations and to other 3^rd parties, including patients.

Healthcare providers have been using direct secure messaging for care coordinating for a while but there may be ways to use it more fully to reduce readmissions, reduce unnecessary testing and procedures and even increase provider productivity. Some benefits may include:

1. Providers can insert secure messages into any point of their workflow.
2. Primary care providers can interact with specialists to potentially result in fewer referrals; ultimately reducing costs to the healthcare system.
3. Hospitals can improve transitions to nursing facilities and nursing homes can improve the communication with providers to reduce the need for readmissions back to the hospital.
4. Medical records departments can interact with the patient send their records more easily saving them time and money.
5. Researchers can seamlessly interact with patients on release of health sciences information.

Once HIEs are fully implemented, query based networks will provide robust data exchanges, but DSMs will continue to be valuable especially for smaller practices and hospitals that do not have the means to implement sophisticated EHRs.

Perficient places the consumer at the forefront of healthcare technology solutions. Our dedicated national practice, combined with Perficient’s more than 1,000 business and technology professionals, delivers innovative and intelligent solutions for hospitals and health systems, integrated delivery networks, health plans, life sciences and state and federal government agencies. Perficient is uniquely positioned to deliver HIE assessments and strategic roadmaps. Stop by our booth at #HIMSS14 booth #2035 to view our case studies.

Follow me @teriemc

See my blog “What is the role of health exchanges to maximize our time in the doctor’s office?” to learn more.

As healthcare organizations prepare for full scale integration of electronic medical records through EHR and enterprise wide data warehouse initiatives, identity resolution is a priority for everyone.

A Master Person Index (MPI) is a solution intended to solve the common problem where multiple systems or applications within the organization gradually become inconsistent with the most current data. When this information changes and only one system is updated, the MPI solution ensures that the change is propagated to all other systems to create the single best view. The MPI may be found at the single system level, facility level, enterprise or health information exchange (HIE) level. A “person” in the healthcare context may be a physician, patient, member, payers, etc.

Data management is one of my favorite subjects and I’m very excited about the evolution of MPIs for identity resolution, as well as, other Master Data Management solutions. But let’s focus on the benefits of the Master Person Index. Some examples are:

• Correctly matching a point of reference to patient, clinician, payer or other entities within and across the organization can improve care coordination and ensure trusted information in an emergency.
• Accuracy in building the “best single view” of the patient will empower both disease management and research activities.
• Using an MPI solution will enhance governance and risk management capabilities since MPI solutions are compliant with IHE audit messaging and recording standards.
• Using an MPI automated solutions provide a means to interpret and analyze patterns in incoming data to help trace lineage of the values to determine original source of bad data.

Perficient’s dedicated national practice, combined with hundreds of business and technology professionals, delivers innovative and intelligent solutions for hospitals and health systems, integrated delivery networks, health plans, life sciences and state and federal government agencies. We are uniquely positioned to deliver identify management assessments and strategic roadmaps. Stop by our booth at #HIMSS14 booth #2035 to view our case studies.

Follow me @teriemc

See my blog “Direct Secure Messaging and Improved Care Coordination” to learn more about the Oracle HIE platform.

What’s transforming the ways in which healthcare is provided?

• legislation
• new competition
• innovative incentives
• a call to refocus on priorities
• a more empowered and digitally engaged consumer, who has more and greater expectations for quality of care and convenience of care.
• a renewed attention on healthcare by the consumer market thanks to ongoing press about healthcare.gov and the Affordable Care Act
• new services and business models in healthcare that we’d never seen before recent reform

…and much more.

Susan DeVore, CEO of our partner and client Premier healthcare alliance, wrote a post yesterday fro HealthAffairs.gov titled, “The Changing Health Care World: Trends To Watch In 2014.” In the article, she introduces the new trends she expects to see in healthcare this year. We are also seeing each of these trends impact conversations about investments our clients need to make this year and next year.

I have summarized the trends below.

1. Investments in Chronic Care –

• Chronic conditions increase costs by 3X, so, “The biggest health care consumers are those with multiple chronic conditions.”
• We should see more investments in Ambulatory ICUs and patient-centered medical homes as providers work to improve their shared savings payments and better manage chronic conditions within primary care facilities.

2. New Job Roles in Healthcare

• DeVore has seen an increase in the popularity of hiring “health coaches” who are there to listen, inspire and motivate and spend time getting to know the patient’s family and life situation as it affects their ability to both access care and to care for themselves.

3. Home Health Care

• Back to the days of the house call.
• “Marketing firm BCC Research predicts that the market for remote monitoring and telemedicine applications will double from $11.6 billion in 2011 to about $27.3 billion in 2016.”
• Technology is increasing access and convenience of healthcare from outside the traditional care setting, especially for rural, hard-to-access locations.

4. Employer Attention on Health

• The Accountable Care Act lets employers increase wellness incentive dollar value from 20 to 30 percent of total coverage.
• Many employers offer incentives for pedometer use or for quitting tobacco or lowering obesity and diabetes rates.

5. Private exchanges become more popular

• This puts the power of choice into the hands of the employee / consumer
• Employees can also customize their coverage to their own needs and better manage their healthcare budget.

6. Further movement toward Value-Based Purchasing

• The Hospital Value Based Purchasing (SVSP) program penalizes provider organizations who fall behind their peers on key metrics such as clinical quality and patient satisfaction, with the penalties set to go up in the next few years. The program also rewards organizations that improve their score over time and outperform industry benchmarks.
• Sustainable growth rate (SGR) is a method for calculating physician payments. Congress will vote this year to “fix” problems with SGR, moving physicians to VBP. This will tie incomes to quality of care and cost improvements.

7. Data will begin to talk as walls fall down

• DeVore says, “Providers are inundated with new technologies that enable them to automate processes and capture new types of clinical data.” That’s quite an understatement from what I have seen.
• More systems will become open to innovation and sharing of data this year, in a provider-led push to do so.

8. New and more creative partnerships in healthcare

• Last year, “drug chains partnering with physician groups to create ACOs based around retail clinics,” says DeVore. But this year, “look for the trend to include community-based groups, including social service agencies, area gyms, and other non-health care service providers.”
• Providers are beginning to look at the “whole person,” and this means unconventional, innovative ways of providing care.

Exciting stuff!

Today, Masachusetts moved into phase 2 of a very important Health Information Exchange (HIE) project. It’s called Mass HIway Health Information Exchange, and it was announced at Beth Israel Deaconess Medical Center, where Healthcare CIO, John D. Halamka works.

John D. Halamka, MD, MS, is Chief Information Officer of Beth Israel Deaconess Medical Center, Chairman of the New England Healthcare Exchange Network (NEHEN), Co-Chair of the HIT Standards Committee, a full Professor at Harvard Medical School, and a practicing Emergency Physician.

Halamka is one of my favorite people to follow in healthcare technology (@jhalamka). He blogs at “Life as a Healthcare CIO” and wrote about the news today in a post titled “The Next Phase of Healthcare Information Exchange.”

Halamka gives us a sense of all of the work that went into this next stage of the project, including:

• a patient index to track which patients have opted into exchanging their patient data with other providers
• a portal for providers to log in and access the data
• an EMR request system to support the request & release workflow
• a system to keep audit trails of all of the activity

For those who are fans of a “No More Clipboards” type of future in their own patient care experience, news like this is very exciting. Halamka talks about how this massive project taking place in Massachusetts is an example of how the future of healthcare information exchange can be convenient, secure, and lead to better care and health outcomes for us all.

He writes:

“I can imagine a day in the next few years, when all patients in the Commonwealth, with their consent, benefit from secure, coordinated care. My mother suffered a major medical error in California because of inaccessible primary care records. I truly believe that my 20 year old daughter, attending Tufts University, will see significant reduction in preventable harm in Massachusetts during her 20’s.”

That’s very inspiring.