Full Interview: Understanding the Pandemic’s Spillover Effects on Non-COVID Patients With Julius Chen
meta name="twitter:card" content="summary_large_image">
By: Makena Binker Cosen (CC ‘21)
I was hoping you could start by describing your background as a researcher.
Thanks so much for reaching out. I really appreciate this opportunity. I’m an assistant professor of health policy and management at the Mailman School at Columbia. This is my second year now. I first started learning about health economics and health policy when the Affordable Care Act was being debated and passed. At the time, I was actually working in state government as a policy advisor. It was just really interesting to see health policy start to take shape, especially at the state level, thinking about how we are going to implement health insurance exchanges or how these pieces of legislation are going to move forward. I saw the role that academic research plays in policy making first-hand. After I finished my time in the State Senate, I started graduate work at the Wharton School. I did my PhD there in applied economics with a focus on health economics. After that, I did a fellowship at Harvard. I was at Harvard Medical School for two years—that’s when I first started thinking about issues of healthcare payment and delivery. I started doing research looking at alternative and value-based payment models, new models for paying physicians to encourage efficiency and cost-reduction while still maintaining the quality of care. For that reason, I became very interested in Medicare and associated innovation around physician payment. That background built my current research at Columbia prior to COVID. If you think about it, the majority of non-elderly adults in the U.S. receive their health insurance through their place of work. In the U.S., we have a system where we’re very dependent on employers for our health insurance. Within this framework, I feel like employers can really play a large role in shaping the costs and the quality of healthcare that we can get, especially in the case of large employers. For instance, take Silicon Valley. Many companies have these really nice on-campus clinics and gyms. You could basically leave your desk, go get primary care, and then very conveniently be back at your desk within an hour. These employer-based innovative models of care delivery are what I began studying with my PhD dissertation. What does that mean for your outcomes? What does that mean for your spending? What does that mean for your utilization of other types of services? For example, employers like Walmart are now offering low-cost insurance plans for their employees where they get all of their care from one provider group, like Emory Healthcare. Another example would be Walmart negotiating with John Hopkins Hospital and saying, “We’re going to fly any employee, regardless of where they are in the country, to your health system if they need spine surgery, and we’re going to cover their travel costs too.” Employers hope this direct contracting model will lead to better outcomes and cost-savings down the line. Broadly, that's what has been on my research agenda. Then, COVID happened. I started to think, “How can health economics, health policy, and applied economics be used to study some of the things that are going on?” Especially in the early days of the pandemic, I felt myself wishing there was something I could do. Then, I started thinking of what we can do with policy or economics to evaluate what is going on and maybe say something about the current state of government response to the pandemic or what’s happening outside of the numbers of cases and deaths.
Great, that brings us up to the next question. What has your transition into COVID research been like? What have you been able to bring to the table and how has that shaped the objectives of your research project right now?
Back in mid-March, the first thing that stood out to me about COVID was school closures. With schools closing, a lot of students would not be able to get school lunches, for example. We didn’t know how long this was going to go on for. If school age children are not able to go to school, that interrupts their educational process and ability to get needed nutrition. What will be the lasting impacts on their health, nutrition, chronic condition management, and ability to find a job and remain employed? There are a lot of indirect effects of the pandemic that we will need to pay attention to down the line. After talking with some of the senior faculty in my department, I started thinking that there are a lot of interesting indirect effects within healthcare that we don’t understand. Towards the end of March and the beginning of April, things started to really worsen with COVID, especially in New York City. I read many articles and news clips where people were talking about how a lot of elective procedures are being cancelled and in-person visits are being postponed. People who used to be getting primary care are afraid of going to their offices and being exposed to COVID. This is a sudden shock to demand for healthcare. What might be the consequences of people not being able to get some types of treatments? At first, I thought that maybe these are just elective procedures that can wait, like hip replacements. Then I started seeing articles where patients were saying, “I need an organ transplant. I need a surgery to remove pre-cancerous tissue. I need to go in for regular cancer treatment.” That’s how I started thinking that there was an interesting project to do here that looks at some of the consequences or the potential effects of delayed or forgone care on non-COVID patients. The news will tell you the number of confirmed cases and deaths. Even then, researchers believe those numbers are underestimates. If you think of the full toll of the pandemic, there are a lot of indirect effects that we don’t understand as well. Not just on mental health, the economy, or education, but on how the healthcare system is functioning; on non-COVID patients who also have urgent needs. Heart attack, strokes, and cancer are still happening in the background. That’s where I launched into thinking about the indirect effects on non-COVID patients' outcomes.
What are some of the specific questions or themes you are looking at? How do you hope to address them in this project?
I envision two main aims. The first aim would be to look at the potential effects of delayed or forgone treatment on non-COVID patients’ short- and long-run health outcomes. For example, two groups I would be interested in studying are cancer and transplant patients. How long does it take them to get their procedures done? How do their outcomes compare to patients who did not have their treatments delayed? COVID is starting to be on the decline in many cities, so a lot of health systems are trying to bring back these procedures. This provides a very interesting setting to study what happens during this delay.
The second aim involves studying what is happening to in-person primary care utilization. A study from The Commonwealth Fund finds a large drop in in-person ambulatory care utilization that occurred during COVID, followed by a slight rebound. As the months progress, how much of a rebound will we see? Will things get back to pre-COVID levels? How long will it take for people to start using in-person care again? It seems that it’s definitely ramping back up and people are going back to see their physicians and other types of clinicians. On a similar note, there are New York Times and ProPublica articles where people are saying that there are not that many heart attacks, strokes, or appendicitis cases coming into hospitals anymore. That doesn’t make sense. Those types of conditions are still happening. It’s just strange that they wouldn’t be showing up in the hospital. A lot of these articles claim anecdotal evidence, but there is no supporting data available yet. That’s one thing I’m very interested in trying to figure out. Can we find data that shows what’s happening with these patients? Are people with heart attack, stroke, and appendicitis just not going to the hospital? Are they dying at home? Are they waiting for their condition to be very serious? And if they wait until their appendix has ruptured already and now have a very complicated and severe condition, would it have been more easily handled if they got treatment earlier?
That’s a very good point. It’s definitely something I’ve also wondered with all of these articles coming out. It’s really good to see someone trying to find those answers. You’ve described interest in several stages of the pandemic, including pre-pandemic times, its early beginnings, months in, and — what will hopefully come soon — a post-pandemic world. I’m curious to know, what general timeline do you expect for this project?
That’s a good question. Right now, I’m in the process of working with a Master of Public Health student in the Mailman School to figure out what data sources are available. This has been one of the biggest challenges. The pandemic and its effects are moving so quickly, and they keep evolving. There is a lag between when data is collected and when it becomes available to researchers. To look at health outcomes among some of these non-COVID patients, we could use electronic health record data from hospitals. The only issue is the approval process for getting that data can be a bit lengthy. However, once we are able to find a willing partner to provide health record data, analyzing it and getting preliminary results may only take a month or two. At least right now, we are not going to apply really fancy econometric methodology to do this work. A lot of the initial work, and some of what will be the most important work, is just looking at raw trends in the data. For example, what’s going on with utilization of care?
Another component of my planned research, and the New York Times has been talking about this, is examining the issue of excess mortality. For example, you could use historical data to generate a projection of deaths in March 2020 for New York City. It turns out that the actual number of deaths is much higher than the estimated number of deaths. Roughly 80% of these excess deaths are attributable to COVID cases or probable COVID cases. I would like to figure out what is going on with the other 20% of excess deaths that are not attributable to COVID. Again, it has been tough to find data for that, but the CDC is one potential source. There’s also some data available through the NYC Department of Health and Mental Hygiene. Now, I’m trying to figure out what data is publicly available that we can get and use right away. After we figure that out, within one or two months, it’s possible to start getting some preliminary findings.
That would be great. You mentioned getting data will be a challenge, which would be a huge hurdle for a project that relies on processing data. Are there any other challenges or caveats you expect at the moment?
Yes. The first challenge is any lag in collecting and getting the data. The CDC has done a great job compiling data and the New York Times has actually been utilizing a lot of it for their interactive analyses. If we want to get actionable findings out there quickly, the best bet will be to use publicly available datasets like those from the CDC. The New York Times or FiveThirtyEight could help with data also. The second major challenge will be on the methodological end. How do you isolate the effect of COVID? Say you find some effect on health outcomes for non-COVID patients, how do you know that is because of delayed or foregone treatment and not because they lost their job or health insurance or something else going on in the healthcare system? It is going to be difficult to say that delaying treatment or missing this appointment led to this result, because there is so much going on in terms of mental health issues, or the economy, or losing your job or your partner losing their job. There are all these other confounding factors that could be taking place. It is going to be tricky to isolate a particular causal relationship.
I appreciate how deeply you’ve been thinking about all of this. You’ve outlined some areas in which healthcare has been impacted by the pandemic. What is your perspective on the future regarding your research and the broader impacts of the pandemic?
As for my research interests, these projects that just came onto my radar during the pandemic are really interesting, but they are very time-sensitive. I would like to work on them and try to get these findings out quickly. After that, I would like to pivot back to some of my core interests, such as innovation in healthcare delivery and payment. There is going to be an interesting opportunity to see how employers are responding to the pandemic. There have been so many lost jobs. We’ve seen wide variation in terms of the impact on employers. Employers with employees who can work remotely may have been less impacted than companies or industries where people have to go in to work. Moving forward, employers will need to think a lot about investing in worker health. How do you protect your workforce? How do you prepare for these types of situations where health insurance or healthcare might be disrupted? Hopefully, there’s not another pandemic, but this will challenge employers to think, “We need to be careful of these types of situations where there could be a devastating impact on healthcare and health insurance for employees. So, how do we restructure our benefits and our approach to worker health?” That’s going to be something that I’m very interested in keeping up with. In a post-pandemic world, what types of investments are employers going to make?
As for the broader impacts of the pandemic, two thoughts. First, it’s interesting that we are seeing a rebound in outpatient visits, as shown in recent Commonwealth Fund research. I think it was in March that ambulatory care visits dropped by 60%, compared to pre-COVID levels. Now, they’ve rebounded slightly, but volume is still about 30% below pre-pandemic levels. It will be really interesting to see how long it takes before people start to go back to interacting with healthcare providers on a regular basis. It may take some time, but I do think eventually we will get back to a place where people feel comfortable doing so. People need to see their clinicians and need to get care. Second, COVID has been a big shock to health systems’ finances. There are a lot of hospitals that are reliant on elective procedures for revenue. When those got canceled or postponed, that was a big shock to their finances. Now that visits are starting to rebound, the big question to consider is what are hospitals’ finances going to look like? Are some going to have to close? Are some going to be okay?
Overall, I think the pandemic is going to play a very big role in the 2020 election. I’m very interested to see how the candidates talk about reform in the U.S. healthcare system in light of what happened with COVID. Maybe, we’ll see more momentum towards expanding affordable health insurance coverage options for Americans. It is critical for people to have adequate and stable health insurance coverage, in case something like this happens again.
Most definitely. Moving forward, do you think innovation in terms of employer-based health care will be stifled or transitioned to a different model?
In the United States, we have a system where, in general, adults under the age of 65 are largely dependent on their employers for insurance coverage. With the pandemic, a lot of people who lost their jobs also lost their insurance coverage. I think that this is going to come up in future debates. Do we need to rethink this system where insurance is so heavily tied to the place of work and to your employer? What happens if you lose your job and you lose access to health insurance or healthcare services? I also think that this pandemic will actually encourage employers to invest more heavily in worker health. Employers are seeing how important it is to have a healthy workforce and to have a workforce where people are insured and covered just in case there are any shocks to the economy, like what is happening now. If anything, we now see the importance of innovations that employers are testing to encourage improved quality and efficiency of care, and to restructure the way that employee healthcare is delivered and paid for. I think that these types of innovations will actually pick up momentum in upcoming years.
Under this model of employer-based healthcare, where do self-employed individuals stand?
That’s a good question. The focus is often on people who are employed by mid-to-large-sized corporations. Those corporations have the resources and ability to pull together people to buy insurance plans. It’s much trickier and more expensive for people who are self-employed. I think that’s going to be a major issue coming up in the 2020 election, especially in light of what has been happening with the pandemic. Do we have robust enough options for self-employed individuals to buy affordable health insurance coverage? The Affordable Care Act tried to address this through the creation of health insurance exchanges, but there is still more work to do. What are the mechanisms that we need to have in place to make health insurance coverage more affordable for self-employed or unemployed individuals? It’s going to require broader healthcare reform to address those issues.
Wrapping up, do you have any recommendations for patients who are currently facing delays or cancellations for non-COVID-related procedures and treatments?
I haven’t discussed this directly with any patients, though I have heard cases and read articles about it, such as those on ABC News and ProPublica. Health systems are beginning to ramp up in-person visits and procedures again. Many providers have responded well and adjusted their operations to ensure that the risk of COVID exposure is low, so patients should feel safe to receive in-person care. On another note, while it’s tough right now, maybe there is a possibility that Columbia could collaborate with one of the major health systems in New York, like NewYork–Presbyterian, Mount Sinai or Northwell, to push out a survey or interview patients to see, “Has this issue affected you? Are you afraid to seek care? Have you had your treatment delayed? How do you feel about that?” I think that these results would be really fascinating.
It really would be. Is there anything else you would like to add?
Right now, things are evolving so quickly with the pandemic. There is concern about what is going to happen in terms of re-opening and if we will see flare-ups of COVID again in certain areas. We are really seeing the power of data in all of this. That includes electronic health record data, GPS coordinate data, or even data that Google or Apple could potentially collect. We are fortunate to live in an era with robust data collection capabilities. There is so much power analyzing these large datasets to uncover any relationships that might exist between, say, containment policies or other strategies and outcomes that are of interest to the general public, like access to care, cost of care, or COVID-related outcomes. Especially with the Columbia Science Review getting the word out about what research is going on, it will be interesting to use those analyses to see what did or didn’t work in terms of the responses to coronavirus.