“In the beginning of COVID-19, everything was so urgent, and we just needed to keep so many people from dying,” said Sara Y. Tartof, PhD, MPH, a research scientist in the Division of Epidemiologic Research in Research & Evaluation. “As we got a few years out, we continued to try to prevent deaths, but we also have been able to look at longer-term questions of health care utilization and see the longer-term impact of the virus on people’s health.”

Cheng-Wei “Charlie” Huang, MD, a hospitalist at the Kaiser Permanente Los Angeles Medical Center, saw 2022 as a transitional year in which the COVID-19 hospitalization rates dropped compared to 2020 and 2021.

“We saw, however, a sustained increase in health care utilization relative to the pre-COVID-19 era that was no longer driven by acute COVID-19,” he said. “There are different theories on this increased utilization, whether it may be due to a direct consequence of COVID-19 or a result of delayed care-seeking for non-COVID-19 health care issues as the pandemic is fading, or a combination of both.”

Some of the studies Kaiser Permanente Southern California researchers conducted in 2022 on the aftermath of COVID-19 included the demand that COVID-19 patients put on the health care system after infection, how medications might reduce hospital admissions and readmissions, and new ways to identify patterns of the pandemic.

“COVID-19 can have long-term impacts on health care utilization, and ongoing management of patients’ health care needs is necessary,” said Debbie Malden, DPhil, MSc, epidemiologist and digital surveillance consultant with the Division of Epidemiologic Research in Research & Evaluation. “Health care organizations and patients should have an awareness of the long-term and multi-organ system effects of post-COVID conditions.”

Understanding COVID-19’s impact on health care at Kaiser Permanente Southern California

The first of these studies was published in January 2022 in the Journal of General Internal Medicine.¹ It showed that patients with COVID-19 put high demands on the health care system whether they had been hospitalized with the disease or not.

“During the pandemic, we recognized that there were many complications associated with
COVID-19 infection beyond severe cases requiring hospitalization,” said the lead author, Brian Huang, PhD, MPH, who was then a biostatistician with Research & Evaluation and is now an assistant professor with the University of Southern California Keck School of Medicine. “With the rapidly growing number of infected individuals and the emergence of what is now known as long COVID, we hypothesized that there would be increased utilization and demand for health care resources. At the time, there were many studies on COVID-19-related hospitalization, but there was less information on outpatient and emergency department utilization following COVID-19 infection.”

In this study, researchers assessed the medical records of 64,011 Kaiser Permanente Southern California members who had been diagnosed with COVID-19 from March through September 2020. The average number of visits over 30 days after diagnosis was nearly 7 for patients who were not hospitalized, and approximately 12 for those who were admitted.

Most of the health care visits after diagnosis were virtual telehealth appointments for COVID-related or respiratory-related reasons.

Dr. Brian Huang said that it had been important to characterize how the pandemic, then still in the early phase, was placing demands on all facets of health care utilization.

Understanding this information would help health care organizations anticipate patient needs and appropriately allocate resources to specific departments and patient populations,” he said.

A broader look at health care utilization

Seven months later, a study led by Dr. Tartof was published in JAMA Network Open.² This study broadened the cohort to 8 different health care systems and extended the time period in which patients were followed after COVID to 6 months.

“Our clinicians had put everything they had into prevention and treatment,” Dr. Tartof said. “And we needed to know what to expect as a health care system. As you can imagine, COVID-19 put incredible stress on the health care system.”

The study showed that contracting COVID-19 was associated with a 4% increase in the use of health care services during the 6 months following initial infection.

“This study highlighted the potential for COVID-19 to exert an ongoing demand on health care organizations,” Dr. Tartof said. “A 4% increase in encounters applied across a large population is a large number of visits associated with substantial cost. In this case, it meant over 27,000 extra encounters among the 8 health care organizations included in this study.”

Paxlovid safety and effectiveness studied

Dr. Tartof worked on other studies related to the aftermath of COVID-19, including 2 examining the effectiveness and safety of Paxlovid for patients who had contracted COVID-19.

In December 2021, the Pfizer-BioNTech medication Paxlovid received emergency authorization as a treatment to prevent severe health problems from COVID-19 that could lead to hospitalization and death. But by May 2022, the medication was being linked anecdotally to a rebound in COVID-19 symptoms following recovery.

The U.S. Centers for Disease Control and Prevention reached out to Dr. Tartof. Could the scientists at R&E harness their research knowledge and robust electronic health records to see if there were any serious issues? Within 3 weeks, Dr. Tartof and Dr. Malden were able to quickly look at this question within the KPSC patient population.

“We were able to describe the proportion of those who were admitted to the hospital or visited the emergency department for COVID-19-related illness after being treated with Paxlovid,” said Dr.
Malden. “Our analysis found that less than 1% of patients who received Paxlovid ended up in the hospital or emergency department for COVID-19 in the 5 to 15 days following treatment.”

The study was published in June 2022 in the CDC’s Morbidity and Mortality Weekly Report.³

Dr. Tartof went on to also research the effectiveness of Paxlovid in a study that was published in The Lancet Infectious Diseases4 early in 2023. That study showed that receiving Paxlovid within 5 days of the start of COVID-19 symptoms was associated with substantial reductions in the risk of hospital admission or death, she said.

Clinicians also examined COVID-19 aftermath

Dr. Charlie Huang led a study of Kaiser Permanente Southern California members who were hospitalized for COVID-19 that identified disease-specific factors associated with lower chances of readmission.

The study was published in September 2022 in the Journal of General Internal Medicine.⁵ The findings added to a growing body of research that helped physicians more confidently decide on the best and safest timing for discharging a patient.

“The study was from early in the pandemic, 2020, and found relatively low rates of readmissions; identified the protective benefits of remdesivir treatment prior to discharge; and recognized
the importance of symptom duration when determining COVID-19 care,” Dr. Charlie Huang said.

“These factors can be helpful particularly when there is uncertainty regarding the safety of a COVID-19 patient discharge, and providers can feel even more reassured in today’s practice, as overall disease severity has declined since 2020.”

Among the other research Dr. Charlie Huang led, one study looked at whether dexamethasone could help prevent deaths of COVID-19 patients if they were prescribed the medication after their release from the hospital. The findings suggested that dexamethasone should not be routinely prescribed beyond discharge for people with COVID-19. The study was published in JAMA Network Open⁶ in March 2022.

Moving forward with COVID-19 research

Dr. Malden suggested that one promising direction for future research in the aftermath of COVID-19
is to continue developing and refining Artificial Intelligence (AI) technologies such as natural language processing (NLP) to improve disease identification and characterization, and to estimate the burden of the disease.

She led a study that was published in December 2022 in JMIR Public Health and Surveillance.⁷
The findings demonstrated that by “enabling the extraction and analysis of large volumes of electronic health record data, NLP and other AI techniques could help researchers and health care professionals identify emerging patterns of disease,” she said.

Dr. Malden added, “This could enhance our ability to prepare for and respond to future disease outbreaks.”

¹Huang BZ et al. J Gen Intern Med. 2022; 37:830-837.
²Tartof S et al. JAMA Netw Open. 2022;5(8):e2225657.
³Malden D et al. MMWR Morb Mortal Wkly Rep. 2022;71(25): 830-833.
⁴Lewnard J et al. Lancet Infect Dis. 2023 March 15. (Epub).
⁵Huang C et al. J Gen Intern Med. 2022;37:3973-78.
⁶Huang C et al. JAMA Netw Open. 2022;5(3):e221455.
⁷Malden C et al. JMIR Public Health Surveill. 2022 Dec 30;8(12):e41529.