Tracking COVID-19’s Effect on the Nation’s High Blood Pressure
Deploying Livongo’s national dataset to inform public health response
By Bimal Shah, MD
Over the past months, the COVID-19 pandemic and accompanying economic uncertainty have led to an unprecedented rise in stress, anxiety, and other behavioral health issues as we all know. In a national survey fielded by the CDC in June of this year, a staggering 40% of Americans reported experiencing at least one adverse mental or behavioral health condition stemming from some combination of health concerns, social isolation, and loss of jobs and income. Symptoms of anxiety disorder are three times what they were in the second quarter of last year, and prevalence of depressive disorder is four times what it was. Racial/ethnic minorities, essential workers, and unpaid adult caregivers are bearing a disproportionately heavy burden through this crisis, reporting worse mental health outcomes across the board.
Consistent with this trend, at Livongo we witnessed a 136% increase in utilization of our myStrength behavioral health solution in the period from January to May as people have found themselves struggling and in need of more support at a moment when they are cut off from family, friends, and reliable coping mechanisms.
Against this backdrop, we wanted to understand what effect the pandemic and its residual behavioral health impacts have had on Livongo Members living with hypertension. While anxiety and stress do not directly cause long-term hypertension, episodes of psycho-social stress and anxiety are well known to cause dramatic spikes in blood pressure (BP). High levels of stress and anxiety also drive people to engage in unhealthy behaviors such as smoking, overeating, and drinking alcohol that can lead to elevated BP. We also know that the shelter at home orders dramatically decreased people’s activity levels. The combination of all of these factors could negatively impact BP both at an individual and population level.
Livongo for Hypertension Members regularly measure their BP with a monitor and cuff that connect wirelessly with Livongo’s smartphone app. Readings are automatically transmitted to our Applied Health Signals platform where participants can view results and get tips on managing their blood pressure by engaging in healthy behaviors. They can also schedule coaching sessions with a Livongo health coach, share data with healthcare providers, and more. Having all of these interactions taking place on a single digital platform offers us an aggregate view of Member behavior and a unique and representative dataset to track health indicators at the national and regional level.
For this analysis, we looked at the proportion of Livongo Members who had high BP before and during the pandemic and subsequent shelter in place ordinances (covering the date range of Sept 15, 2019 through Aug 11, 2020). We defined high BP as a systolic blood pressure of 130 mmHg or higher or a diastolic blood pressure of 80 mmHg or higher, and we grouped BP values into weekly averages to smooth out day-to-day variation. To provide an ongoing interactive view of these findings, we have designed and launched a National Blood Pressure Tracker on our website which will be refreshed and monitored regularly over the coming months.
As the tracker shows, until late January of this year the percentage of our Members nationally with high BP in any given week was on average 62%. At the end of January, however, we saw a rise in high BP roughly corresponding with the announcement of the first confirmed case of COVID-19 in the US (Jan 21) and the first mass quarantine of residents in the Chinese city of Wuhan (Jan 23). From that point forward, the proportion of our Members with high BP has mostly remained at a heightened level above 62%. An initial analysis of data at the state level shows that in 30 states the percentage of Members with high BP has increased between January and August.
According to the data, high BP reached another significant peak in early to mid April when 68% of our Members nationally registered high BP. This timing was notably following widespread state-based lockdowns (median lockdown date of Mar 23 was derived from this and this analysis). Peak high BP also coincided with the initial release of then historic (and long since surpassed) unemployment figures (April 2), the CDC’s first recommendations to wear masks in public (April 3), and the first peak in virus cases in New York City (April 10) and other early hotspots. While proportions of high BP have gradually lowered since that peak, they remain elevated relative to pre-pandemic levels.
Although this data can’t support a direct cause-and-effect relationship between this specific sequence of events and high BP trends, it does reveal a remarkably clear correlation, underscoring the interconnectedness of behavioral and physical health and the extent to which the pandemic has compromised downstream health and wellbeing for so many Americans.
Among other takeaways, our study provides a stark reminder of the heightened vulnerability of people living with chronic conditions and struggling through this crisis to manage hypertension and other chronic conditions. While no more susceptible to COVID-19, this population is more at risk of serious illness, hospitalization, and death from the virus. As our BP data reveals, the stress and anxiety and social isolation we have all experienced has had an outsized and measurable impact on this at-risk population. Throughout the crisis, at Livongo we have doubled down on support for our more than 410,000 Members to ensure they have access to up-to-date resources, targeted Health Nudges, social and behavioral health support, and 24/7 digital and live coaching — all from the safety of their homes.
By providing near-real-time visibility into the pandemic’s impact on hypertension — a major risk-driver for heart disease and stroke, two of the leading causes of death for Americans — our analysis also speaks to the promise of consumer centered virtual care for the future of public health surveillance, whether in a time of pandemic or not. As Tom Frieden, a former director of the CDC, recently noted, the US response to the pandemic has been hampered by having “both a glut of data and a scarcity of information.” Although he was speaking specifically about tracking COVID-19, it applies equally to the broader collateral damage on other health indicators that we are just beginning to account for. Equipped with data that can inform and elevate situational awareness, we can in the future plan and deploy resources more quickly and effectively and tailor and target responses more precisely.
For our next phase of analysis, which I will be reporting on in future posts, we will be digging into our state- and county-level data on high BP. As we all know, the pandemic’s timing and impact have varied significantly by location. In the absence of a unified national strategy, states — and in some cases municipalities — have implemented their own policies on everything from the mask-wearing to the nature and timing of lockdown and re-opening, instilling varying degrees of reassurance and confidence in their citizens. We hope our regional analysis will reveal additional insights into BP trends that we can continue to use to fine-tune our AI+AI engine and offer Members personalized geo-targeted Health Nudges to help them stay safe and healthy.
In addition, this data will reveal insights into the ongoing impacts of the pandemic and the role of regional policies in mitigating downstream health indicators. Moving forward, our goal is to use these findings to help inform policymakers, employers, and health plans on how to better address overall health through challenging times.