Expert Perspective: PCCI CIO testifies at Texas House Committee in favor of a making it easier to enroll MCO members into electronic communication

Vikas Chowdhry, PCCI’s CIO testified at the Texas State Capitol in front of the TX House Human Services Committee (chaired by Rep. James Frank) in support of House Bill 4343 sponsored by Representative Toni Rose (District 110). This bill will require HHSC to gather member contact preferences and gain their informed consent on the application, and then pass that consent to the Managed Care Organization (MCO), thus making it easier for MCOs to contact their members via their preferred electronic modes of communications such as texting.

PCCI has observed health outcomes greatly improve for Parkland Health Plan’s members through our pediatric asthma and pre-term birth prevention text messaging programs and we fully support this bill that will allow MCOs to expand programs like this.

 

 

 

 

 

 

 

 

(Rep. Toni Rose (center) with Kay Ghahremani, CEO of Texas Association of Community Health Plans and Vikas Chowdhry, CIO at PCCI)

 

Full text of testimony is follows:

HOUSE BILL 4343 (ROSE)

TESTIMONY OF VIKAS CHOWDHRY

Mr. Chairman, members, good morning (afternoon), my name is Vikas Chowdhry.  I am the Chief Analytics and Information Officer for the Parkland Center for Clinical Innovation – PCCI – and I am testifying in support of House Bill 4343.

PCCI started as a department within Parkland Health and Hospital System and was spun out as an independent, not-for-profit organization in 2012 to not only serve the needs of Parkland, but to also pursue additional transformative initiatives that could have a broader impact. PCCI remains tightly connected to Parkland, the Parkland Foundation and the Parkland Community Health Plan. Our collaborative work focuses on the needs of vulnerable populations across North Texas and beyond. Our work focuses on cutting edge uses of AI, Non Medical Drivers of Health, and clinical expertise across clinical and community settings.

Representative Rose, thank you for filing this bill.  Simply put, when a health plan is able to text its members, health outcomes improve.  They can do this now, but (as you heard) MCOs are required to first call a member and ask for their consent.  The problem is not having to seek the consent, which is entirely appropriate, but having to get the consent via a telephone call, because of course most people do not answer calls from numbers they don’t recognize.  In Representative Rose’s committee substitute, HHSC is tasked with gathering member contact preferences and sending those preferences to the MCO.  This should make it more likely that an MCO can contact a member via text.

Why is this beneficial?  At PCCI, we have documented the positive impacts of programs involving the texting of members.  In our Asthma program, we studied Parkland Community Health Plan members under the age of 18 who were enrolled in our text messaging program.  Those in the study received 3-5 educational or reminder text messages per week, including 2-item Asthma symptoms surveys once or twice a week.  We saw a 22.5% increase in asthma outpatient visits, and a six-fold drop in asthma-related visits to the emergency department.  84% of the respondents to our satisfaction survey said that the program has taught them to take better care of their child’s asthma

We performed similar research in our Preterm Birth Prevention Program.  Pregnant women enrolled in the program received four to five messages per week.This study produced positive results as well, with a 24 percent increase in prenatal visit attendance, and a 27 percent drop in preterm birth rates. More than two-thirds of the respondents to our satisfaction survey said that as a result of this program, they feel better prepared to take care of themselves and their baby.

The analysis that we have shared above is based on pre-COVID data. Through COVID period, we observed an overall reduced impact on outcomes across all these groups but even within that period, members enrolled in text messaging program still had better outcomes compared to those not enrolled. So, just through something as simple as a few text messages, we saw better health outcomes and less expensive care.  Texting works.

Mr. Chairman, Representative Rose, members, thank you for the opportunity to visit with you today.  I’m happy to answer any questions.

===========================================================

 

 

PCCI’s Vulnerability Index Shows Lowest COVID-19 Infection Risk Level for Dallas County

DALLAS – Towards the end of March, Parkland Center for Clinical Innovation’s COVID-19 Vulnerability Index has recorded the lowest infection risk since the Vulnerability Index launched in June of last year.

“After the holidays, we had vulnerability index ratings at nearly 200, which meant the COVID-19 virus was running rampant through our community,” said George “Holt” Oliver, MD, Vice President of Clinical Informatics at PCCI. “It is a great relief to see that the highest vulnerability index rating now is only 16.91. This is a triumph for our county’s public health leaders, providers and residents who have made the sacrifices and efforts needed to bend the curve.”

One of the hardest hit ZIP Codes during the past year, 75211, which includes the areas around Cockrell Hill and Oak Cliff, saw its vulnerability risk hit the high of 196.9 in January. This was the highest level any ZIP code in Dallas County reached. By mid-March, its vulnerability rating was 8.74, a dramatic improvement for an area facing some of the most sever socioeconomic issues.

“This is very good news for the residents of the 75211 ZIP code; however, we advise caution going forward,” said Dr. Oliver. “I believe that our new normal will be continued vigilance. To keep COVID-19 from resurging, everyone who can be vaccinated should seek it, and adhere to local health official guidance that includes direction on social distancing and face covering.”

Launched in June 2020, PCCI’s Vulnerability Index identifies communities at risk by examining comorbidity rates, including chronic illnesses such as hypertension, cancer, diabetes and heart disease; areas with a high density of populations over the age of 65; and increased social deprivation such as lack of access to food, medicine, employment and transportation. These factors are combined with dynamic mobility rates and confirmed COVID-19 cases where a vulnerability index value is scaled relative to July 2020’s COVID-19 peak value. The PCCI COVID-19 Vulnerability Index can be found on its COVID-19 Hub for Dallas County at: https://covid-analytics-pccinnovation.hub.arcgis.com/.

Currently, the 75150 ZIP code, at the intersection of Interstate Highway 30 and 635 has the highest COVID-19 risk at 16.91, down from a high of 107.30 in January. The ZIP code 75204, in east downtown Dallas, has the second highest vulnerability level at 15.81, down from a high of 126.5 in January.

PCCI’s forecast of Dallas County reaching COVID-19 herd immunity is still on-track but reaching that threshold is highly dependent residents receiving their vaccinations.

“With vaccinations available to all adults, we need to get in line and get immunized,” said Dr. Steve Miff, PCCI President and CEO. “We don’t want another year to go by where grandparents can’t hug their grandchildren. We have seen how safe and effective the current vaccines are, so it is the responsible thing to do for our friends, families and co-workers to get immunized.”

While always concerning when adverse reactions emerge, the action by the FDA to pause the J&J vaccine is out of “abundance of caution” and it’s a strong signal of how responsive they are to any potential safety concerns. Cerebral venous sinus thrombosis (CVST) with J&J vaccine has been reported in 6 young women (ages 18-48) among 6.8 million doses in the US. To date, Dallas County has administered 61% Pfizer, 35% Moderna, and 4% J&J. The syndrome has been dubbed vaccine-induced immune thrombotic thrombocytopenia.(VITT), based on a similar syndrome after the commonly-used medication heparin abbreviated HITT. The reported rates are much lower than IV Heparin which is used frequently in the hospital. While the risk benefit ratio of continuing to use J& J vaccine in the US COVID-19 vaccination plan may still make sense given the observed case fatality rate of 1.8% of COVID-19, prudence to understand the situation given the FDA emergency use authorization for use is warranted..

The FDA pause for the J&J vaccine will not significantly impact the PCCI initial estimate for Dallas County’s path to herd immunity by June. We were progressing towards herd immunity at a rate of approximately 3% per week, which was ahead of initial predictions. While the allocations for J&J were scheduled to increase and the latest developments will pause those vaccinations likely for days, up to several weeks, we forecast that Dallas county will continue to make progress at 2-2.5% per  week, which maintains the pace for mid-June.

A year in retrospective
With the COVID-19 pandemic ongoing for over a year, PCCI identified the zip codes with the highest average vulnerability from July 2020 through March 2021. These represent areas which have faced the highest risk during the COVID-19 pandemic to date.

Data Sources:
To build Vulnerability Index, PCCI relied on data from Parkland Health & Hospital System, Dallas County Health and Human Services Department, the Dallas-Fort Worth Hospital Council, U.S. Census, and SafeGraph.

About Parkland Center for Clinical Innovation
Parkland Center for Clinical Innovation (PCCI) is an independent, not-for-profit, healthcare intelligence organization affiliated with Parkland Health & Hospital System. PCCI leverages clinical expertise, data science and Non Medical Drivers of Health to address the needs of vulnerable populations. We believe that data, done right, has the power to galvanize communities, inform leaders, and empower people.

###

Authors

Steve Miff, PhD., President & CEO of PCCI, George “Holt” Oliver, MD, Vice President of Clinical Informatics at PCCI and  Thomas Roderick, PhD, Senior Director of Data and Applied Sciences at PCCI.

“BUILDING CONNECTED COMMUNITIES OF CARE” BOOK EXCERPT Case Study – Engaging Patients—Location and Relationships Matter

Following is an excerpt from PCCI’s book, “Building Connected Communities of Care: The Playbook For Streamlining Effective Coordination Between Medical And Community-Based Organizations.” This is a practical how-to guide for clinical, community, and government, population health leaders interested in building connected clinical-community (CCC) services.

This section is from Chapter 6, “Clinical Providers Track.” The purpose of the Clinical Providers Track is to set out the stakeholders and processes required to integrate clinical entities, insights, programs, interventions, strategies, and measurement for the CCC.

PCCI and its partner Healthbox, offers readiness assessments as a service. If you and your organization are interested, go here for more information: https://pcci1.wpengine.com/connected-communities-of-care/.

###

Case Study: Engaging Patients—Location and Relationships Matter

As part of our CCC history, PCCI has developed and tested a number of approaches to identifying individuals within the population of vulnerable and under-served Parkland patients who could benefit from screening for health related social determinants, engaging them in the completion of a brief risk assessment and subsequent linkage to available community resources. As with many of the elements of the CCC, this proved to be a learning experience in which initial, more conventional approaches gave way to new and more innovative approaches of engaging this population to optimize goal
attainment.

RECRUITMENT
Much of the initial work began with screening in the outpatient setting. Parkland has 12 Community-Oriented Primary Care (COPC) clinics located throughout Dallas County to serve local residents. Because the COPCs see a large number of patients on a daily basis, many of whom are considered vulnerable and underserved, these COPCs were determined to be a great
location to conduct the social determinant risk assessments. When a patient checked in for a visit, the office staff would provide the patient with a paper-based screening tool to self-administer. Trained community health workers were available in the waiting area to help, if required. Initially we felt like this approach made sense since the large number of COPC patients translated into large numbers of completed screening surveys. However, while there were a large number of initial screenings, the number was very low of patients that agreed to engage with a PCCI community health worker to connect with local community services. Many stated they were not interested or needed to leave the facility for another commitment. Other patients completed the needs assessment but left the COPC before staff members were able to connect with them. Of these, very few responded to follow-up phone outreach and the ones that did were hesitant about referral to community-based services. The team attributed this gap to the lack of personal engagement at the point of initial screening.

As a result of this initial experience, the team made some changes to the screening protocols. Three concurrent workflows focusing on different points of patient encounters were designed and tested. The three new points included: (1) engagement while the individual was in the ED, (2) engagement of individuals that had already left the ED, and (3) engagement of hospitalized patients on the medical/surgical floors of the hospital.

For the direct engagement while the individual was in the ED, licensed social workers conducted initial face-to-face screenings with patients awaiting care. The social workers were provided a list of eligible patients (those with multiple ED visits in the past year) and went room to room to conduct the screenings and determine if the patients were interested in connecting with community resources. Because many of these patient interactions took place while the individual was in the middle of an ED care visit, the PCCI team member was mindful of this and stepped aside, as needed, to ensure they didn’t interrupt the patient’s care. For those individuals that left the ED before screening, the PCCI team placed these individuals’ names and contact numbers on a sheet and later reached out to them by phone to explain the program and ask if they were interested in receiving information on community resources.

Finally, for those individuals undergoing an inpatient stay in the hospital, PCCI personnel obtained census data reports with information about eligible patients and then staff visited these patients in their rooms to conduct one-on-one conversations to implement the screening tool and to determine if the patients were interested in receiving more information about navigation services to community resources.


As shown in Table 6.1, a key learning from this undertaking was that the site matters in conducting the screenings and successfully connecting people to local programs for support. We learned that engaging patients during their inpatient stay was the optimal care setting in which to conduct screenings and then connect those patients to the appropriate community resources.

Establishing trust with patients early in the process was essential, both for completing the initial screening tool and for facilitating connection to community services. During our initial approach, we relied on self-administered screenings that provided little in the way of opportunity to establish a relationship with patients. Our modified workflow allowed our social workers and community health workers to verbally administer the screening tool and provide additional explanations as part of that exchange. This process also made the transition to navigation services virtually seamless and much more
effective. Feedback from patients has also been positive; most indicated that the information received was useful and many said they would share this information with other family members and close friends.

THE SCREENING PROCESS
The PCCI community engagement team consisted of six community health workers and two master’s-level, licensed social workers. Initially, the team consisted entirely of social workers, but our experience taught us that a blended staff model was more cost-effective. PCCI physician leaders coached all team members on how to be flexible and professional when working in the ED, where care moves at a rapid pace. The team needed to take cues from medical staff on where and when to step in to conduct the screenings. Similar trainings were delivered to those staff visiting patients in the hospital.

Over the course of the 6-month pilot, we were also able to identify a number of key elements that increased both the effectiveness and efficiency of the screening process. For example, we learned that it took on average 15 minutes to complete the assessment tool when it was facilitated by a team member but only 10 minutes when self-administered. While the self-administered survey took less time to complete, we found a much higher percentage of incomplete and inaccurate responses, making many of the screens useless. As would be expected, we also found that older patients—those 65 or older—took on average 20 minutes to complete the facilitated screening survey while younger individuals completed it in half the time. The difference was attributable to the amount of questions asked and attendant conversations, which were much more prevalent with older patients. Finally, once we began to work more closely with the patients and they developed a better sense of the purpose of the work, we encountered very few issues with obtaining consent from the patients to share their information with others.

###

 

COVID-19 fatalities become the leading cause of death in Dallas County one year into the pandemic 

Dallas – With the anniversary of Dallas County’s first COVID-19 death having recently passed, mortalities due to the pandemic has become the leading cause of death among county residents, surpassing heart disease, cancer and strokes in the past year.

According to the Centers for Disease Control and Prevention (CDC), the first death in Dallas County was recorded on March 19, 2020. By March 21, 2021, deaths in Dallas County from COVID-19 stood at 3,763. This surpassed estimated deaths due to heart disease (3,668), cancer (3,356) and strokes (1,015) during that same period.

COVID-19 deaths in Dallas County saw their steepest increases starting in December. On Dec. 21, 2020, deaths due to COVID-19 stood at 1,841, but in the following three months deaths more than doubled, adding 1,922 more casualties.

“This is a sad milestone for Dallas County,” said Vikas Chowdhry, MBA, Chief Analytics and Information Officer at PCCI. “We can see that COVID-19 claimed the most lives following social gatherings and holiday travel beginning with Thanksgiving through Christmas and New Year’s. Starting in December we saw a startling spike of deaths due to COVID-19 that represented more than all of the deaths in the previous months we had experienced during the pandemic. This offers a valuable lesson going forward, that we must remain vigilant to protect ourselves, our families and friends.”

PCCI recently forecast that Dallas County may reach COVID-19 herd immunity by mid-June. However, in order to reach this threshold residents of Dallas County need to continue their efforts to protect themselves from infection. “We are remaining optimistic that we can reach herd immunity by the early summer, but the key is ongoing vigilance, including continued adhering to local health official guidance, social distancing, face covering, and registering for vaccinations as soon as possible,” said Chowdhry.

An animated graphic showing the evolution of the COVID-19 mortality rate in Dallas County can be viewed at https://covid-analytics-pccinnovation.hub.arcgis.com/, PCCI’s COVID-19 Hub for the region. This shows total COVID-19 deaths by day, based on data provided by the New York Times COVID-19 data tracking project. The mortality data includes both confirmed cases, based on laboratory testing and probable cases, based on specific criteria for symptom and exposure. This is per guidance form the Council of State and Territorial Epidemiologists.*

To help protect Dallas County residents, PCCI recently launched the MyPCI App, a web-based program to help inform the residents of Dallas County to their individual risks. The MyPCI App, free to register and use, is a secure, cloud-based tool that doesn’t require personal health information and doesn’t track an individual’s mobile phone data. Instead, it is a sophisticated machine learning algorithm, geomapping and hot-spotting technology that uses daily updated data from the Dallas County Health and Human Services (DCHHS) on confirmed positive COVID-19 cases and the population density in a given neighborhood. Based on density and distances to those nearby who are infected, the MyPCI App generates a dynamic personal risk score.

To use the MyPCI App, go to, https://pcci1.wpengine.com/mypci/, click on the link and register (Using code: GP-7xI6QT). Registration includes a request for individual location information that will be used only for generating a risk assessment, never shared. Once registered, simply login daily and a COVID-19 personal risk level score will be provided along with information to help individuals make informed decisions about how to manage their risk.

About Parkland Center for Clinical Innovation

Parkland Center for Clinical Innovation (PCCI) is an independent, not-for-profit, healthcare intelligence organization affiliated with Parkland Health & Hospital System. PCCI leverages clinical expertise, data science and Non Medical Drivers of Health to address the needs of vulnerable populations. We believe that data, done right, has the power to galvanize communities, inform leaders, and empower people.

 

###

 

*The tallies reported here include probable and confirmed cases and deaths. Confirmed cases and deaths, which are widely considered to be an undercount of the true toll, are counts of individuals whose coronavirus infections were confirmed by a molecular laboratory test. Probable cases and deaths count individuals who meet criteria for other types of testing, symptoms and exposure, as developed by national and local governments.

Steve Miff Webcast on Artificial Intelligence for Non Medical Drivers of Health

Recently, PCCI’s CEO & President, Steve Miff participated on a webcast hosted by the University of South Florida for the Florida Medicaid Drug Therapy Management Program for Behavioral Health. His presentation “Artificial Intelligence for NMDOH” discussed how Non Medical Drivers of Health and AI can work together to help improve care for under-served populations. Click on the image below to see the video webcast:

 

 

 

Get Your COPY OF THE BOOK THAT WILL HELP BUILD SUSTAINABLE CONNECTED COMMUNITIES OF CARE

Available now is a new book from PCCI: “Building Connected Communities of Care: The Playbook For Streamlining Effective Coordination Between Medical And Community-Based Organizations.” This is practical how-to guide for clinical, community, and government leaders interested in building connected clinical-community services.

The book shows how to facilitate cross-sector care coordination; enable community care partners to better provide targeted services; reduce duplication of services across partnering organizations and help to bridge service gaps in the currently fragmented system.

Published by HIMSS, the book will be available on March 9. Reserve your copy today at: HIMSS publishing: https://lnkd.in/eiB2Jq7 Or Amazon.com: https://lnkd.in/eBKH9h7

 

Texas Medicine Magazine highlights success of one clinic’s allergy and asthma pilot program

The September issue of the Texas Medical Association’s magazine, Texas Medicine Magazine, featured the efforts of C. Turner Lewis, III, MD, Medical Director of Children’s Medical Clinics of East Texas, to mitigate the harmful effects of pediatric asthma and alergies. Dr. Lewis employed a pilot program that included elements of PCCI’s predictive modeling to help reduce emergency department visits to zero over the course of a two-month period.

Click on the image below to read the entire article:

 

VIDEO: PCCI’s Women in Data Science & Technology Internship Delivered Immersive Experience

A video from Parkland Center for Clinical Innovation (PCCI) highlights its Women in Data Science and Technology Summer Internship program, with members of the program sharing their valuable experiences.

PCCI’s 2019 summer intern program is made up of area students from Dallas Independent School District high schools, SMU’s Statistics Department as well as students from the University of Texas at Dallas and Creighton University.

This internship program has become one of the most prestigious internship programs in North Texas with a mission to expand opportunities for women in an industry that significantly lacks gender diversity.

 

 

 

Deep Learning Model to Predict Pediatric Asthma Emergency Department Visits

Pediatric asthma is the most prevalent chronic childhood illness, afflicting about 6.2 million children in the United States. However, asthma could be better managed by identifying and avoiding triggers, educating about medications and proper disease management strategies.

Parkland Center for Clinical Innovation (PCCI) has been working with the Parkland Community Health Plan (PCHP) for the past four years to help them with timely identification of high risk asthma members in the pediatric population. We have deployed a logistic regression model on claims data that is the foundation for an innovative care redesign and text based patient engagement program that has show consistent cost savings and clinical outcomes improvement for PCHP members. You can learn more about this program here.

Thus, this issue for us is not merely an academic exercise. We realize that a predictive model with improved statistical performance can yield even better health outcomes for these families and improved cost savings for PCHP. Keeping that in mind, we have regularly looked for ways to continue to improve our model that is not just limited to retraining the model but also finding ways to improve data quality and bringing in additional data sets for workflow improvement. More recently, we decided to retrain the model using deep learning techniques — more specifically using an ANN.

Our deep learning model produced an AUC of 0.845 which was only slightly better than the AUC of the current logistic regression model at 0.842.

You can read about details of our work at arXiv here.

Healthcare has high expectations for the level of transparency from machine learning algorithms deployed in a clinical setting. Deep learning models with their relative lack of transparency are not always the best contender for those situations. However, if the improvement in performance of the model for relevant statistical measures is significant enough, then there can be a strong case for deployment of a deep learning model over say, a logistic regression model.

However, in this case, with the exact same data set and initial feature list, ANN model only produced slightly higher statistical classification power than the Lasso logistic regression. This is consistent with the results from some other published research that has compared logistic regression and ANN models in multiple medical data classification tasks. This study further confirmed that the Lasso logistic regression model developed by PCCI in 2015 could produce desirable statistical performance that is non-inferior to deep learning models which are more difficult to interpret. And in order for our predictive models to be deployed and effectively improve patient care, we need to work closely with clinicians to explain predictions in comprehensive and interpretable formats to build trust and transparency with stakeholders.

For future studies, blender algorithms would be tested against other singular models to achieve better statistical performances. We would explore the temporal relationships in claims data using other deep learning models, like Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM).

Parkland Center for Clinical Innovation (PCCI) uses data, advanced analytics and Artificial Intelligence to help individuals lead healthier lives.