Providing care for the 99.9% during the COVID-19 pandemic: How ethics,equity,epidemiology,and cost per QALY inform healthcare policy

Introduction

By July 1st, 2020, 10.5 million people, ∼0.1% of earth’s ¹ population, had experienced a Coronavirus Disease 2019 (COVID-19) infection ² Thus, it is important that healthcare policies, both in hospitals and in the public health domain, balance controlling COVID-19 with provision of healthcare for all people and resumption of commerce. One should not minimize COVID-19’s potential lethality, but we must also remember that 80% of infected people are only mildly symptomatic, <10% require hospitalization, and ∼5.6% of people die. ¹ While ensuring timely healthcare for the 99.9% is not controversial, designing and implementing policies to achieve equitable care access for people with COVID-19 versus other health needs is complex.

The highly infectious nature of the SARS-CoV-2 virus and our lack of natural or vaccine-based immunity create an ethical tension that engages policy-makers, practitioners, and the public differentially depending on their risk tolerance and their understanding of the consequences of policies that focus solely on mitigating COVID-19. Collateral risks of COVID-19-centric policies include excess mortality from deferred care of other diseases and adverse personal and societal economic consequences. Healthcare policy is best framed with the recognition that fear exists, that people have varying health literacy, and that even health literate people have variable tolerance for risk. Gaining buy-in to an ethical policy to deal with COVID-19, while treating all other diseases, requires not only strong epidemiologic underpinnings and sound economic analysis, it will also require clear, coherent, communication and dialogue between governments, healthcare professionals, and the public.

Currently, the global approach to slowing the spread of COVID-19 largely consists of non-pharmacologic interventions including physical distancing, closing schools and nonessential businesses, hand hygiene, masking, the use of quarantine, and testing for the virus. Even these simple interventions may have asymmetrical adverse impact on specific populations (women, minority groups, people of low income). ³ Without explicit attention to ethics and equity, policy focused on pandemic management can inadvertently lead to excess morbidity and mortality from other diseases, exacerbate economic challenges, and widen health disparities. ³

Sound policy must acknowledge geographic and temporal heterogeneity in COVID-19 epidemiology. The incidence and prevalence of COVID-19 vary greatly among countries and even within smaller jurisdictions, like provinces and cities. In Canada, the brunt of COVID-19 is in Ontario and Quebec, with Quebec accounting for 55% of all cases and Montreal accounting for over 25% of the national total. ⁴ On June 30^th, 2020 there was an ∼10-fold variation in COVID-19 prevalence by region, from 41.8 cases/100,000 population in Kingston to 418.8 cases/100,000 population in Toronto. Moreover, within Toronto, there are hot spots with over 1,700 cases/100,000, where more aggressive intervention is justified ⁴ Finally, policy makers must understand non-geographic variables relevant to the pandemic. First among these is the fact that long-term care facilities account for less than 0.5% of Ontario’s population but over 70% of all COVID-19 deaths. ⁵ Governments struggle to create agile and targeted policies. A nuanced approach is required to avoid deferred care by ramping up elective care where it is safe, while focusing on making long-term care facilities safe and intensifying COVID-19 interventions in more afflicted neighbourhoods. ⁴ Ethical and equitable healthcare policy should not exacerbate the financial trauma caused by the pandemic. Economic and health strategies should afford timely access to healthcare, employment, and food security. When making policy regarding the intensity and duration of COVID-19 interventions, we should consider the cost per Quality Life Year (QALY), defined as cost to achieve a year of perfect health. Historically, expensive publicly funded healthcare interventions, such as organ transplantation and haemodialysis, have been evaluated this way.

During the initial stages of a viral pandemic, it is crucial to adhere to rigorous non-pharmacological interventions to flatten the curve and prevent the rapid spread of the virus. This approach does not eliminate the virus, but it can attenuate the disease surge to avoid overwhelming the healthcare system. During the 1918 Spanish influenza pandemic, cities which effectively implemented non-pharmacological interventions, like New York City, had lower mortality rates than cities that did not (eg, Philadelphia and Pittsburgh). ⁶ Likewise, in a study of China, South Korea, France, Iran, Italy, and America, non-pharmacologic interventions prevented or delayed COVID-19 cases that would have otherwise occurred. ⁷ The type of non-pharmacologic interventions deployed varied by country but included travel restrictions, social distancing through cancellations of events and suspensions of educational/commercial/religious activities, and quarantines and lockdowns. These interventions slowed the early exponential phase of the epidemic and averted roughly 530 million total infections, with most of the benefit deriving in China and South Korea, where interventions were most rigorous. ⁷ The conclusion that early non-pharmacological interventions work is supported by the excess mortality rates from COVID-19 in Sweden and the United Kingdom. These countries didn’t aggressively implement non-pharmacological interventions to flatten the curve and had excess mortality relative to other European Union countries. ⁸ However, while non-pharmacologic interventions delay the spread of COVID-19, they are not a cure. Since SARS-CoV-2 is a new contagion, against which we appear to have little immunity, it is unclear whether flattening the first wave of COVID-19 will prevent later secondary and tertiary disease waves. Moreover, 4 months into the epidemic, the consequences of an initial laser focus on COVID-19 care at the expense of all other care are emerging. A cost/QALY approach to our next policy choices could be useful in assessing the consequences on diverse outcomes, such as health of the 99.9% and unemployment, in addition to prevention of COVID-19.

A cost/QALY approach favours strategies and policies that extend healthy life. Diseases like coronary artery disease fare well by a cost/QALY metric since the interventions are reasonably affordable and effective, the recovery period usually brief, and many patients return to employment. In the case of COVID-19, most mortality afflicts the frail and elderly population, especially those with multiple comorbidities. In Ontario, as of June 11, 2020, only 4% (105/2,475) of deaths occurred in people younger than 60 years. ⁹ This death excess in the elderly population occurs despite the fact the infection prevalence is slightly higher in those younger than 60 years (249 cases/100,000) versus those older than 70 years (214 cases/100,000). ⁹

Since the majority of people who are dying are often residents of long-term care facilities, ethical health policy should be targeted here. ⁵ A cost/QALY assessment of the impact of current broad, societal non-pharmacologic interventions on the health of long-term care residents would likely show high cost/QALY. To achieve ethical care for long-term care residents, without depriving the 99.9% of healthcare for other disease and damaging the economy, we need more strategic policies. Failing a cost/QALY metric does not mean one does not intervene; rather, it requires a different intervention. The cost per QALY threshold that is commonly used to define interventions that merit public funding in the United States is $50,000. This historical threshold likely stems from American policy pertaining to public funding of hemodialysis. ¹⁰ With age and increasing numbers of chronic diseases, QALY expectations decline. Jia et al. found the age-adjusted QALYs over 65 years was a mean of 5.8 years for men and 7.8 years for women and was even lower for those with depression or congestive heart failure (1.1-1.5 years for men and 1.5-2.2 years for women). ¹¹ In 2019, 90% of Ontario’s long-term care residents had cognitive impairment, 86% needed extensive help with daily activities, 80% had neurological diseases, 76% had cardiovascular diseases, and 64% had dementia. ¹² In light of the high burden of chronic diseases in residents of Ontario long-term care facilities, the potential to improve QALYs would be low. Since most COVID-19 deaths occur in residents of these facilities, a population-wide restriction on the economy and healthcare systems would also have an unfavourable cost/QALY rating, since most infected people outside long-term care facilities are not at high risk of dying but are at risk from deferred care for other diseases and economic hardships. This might argue for highly targeted long-term care interventions. The logical and cost-effective intervention would be to immediately require all long-term care facilities to offer single rooms for all residents, hire appropriate numbers of properly trained personal support workers, acquire adequate personal protective equipment and document the goals of care for all their residents. To try and reduce COVID-19 deaths by a prolonged shut down of broader society and restriction of access to healthcare for most patients is a misunderstanding of the disease’s true epidemiology and fails the cost/QALY test. Broad and expensive societal interventions, which cannot provide good cost/QALY outcomes for Long Term Care residents, may also exacerbate their isolation and suffering by limiting their freedom to have visitors.

When considering the ethics of healthcare policy, it is important to remember that cancer and cardiovascular disease account for more than 54% of all Canadian deaths. ¹³ An excessive focus on COVID-19 would likely increase excess mortality, defined as the gap between the total number of people who died from these causes in a specified region and time period versus the historical average for that place and time. Because access to elective healthcare for the “99.9%” is currently limited in Canada by government policy, one would expect excess mortality might increase until full care access is resumed. At Kingston Health Sciences Centre, our ambulatory, in-person clinic visits are ∼33% of pre-pandemic levels. To meet our ethical responsibilities to provide equitable care for all people, innovations have occurred. While we continue to see patients requiring non-deferrable care in person in our clinics, we have expanded our use of video visits ¹⁴ and other e-health tools. However, most procedures and many critical patient-doctor interactions must occur in person, and these have been delayed.

As the first wave of COVID-19 declines, excess mortality increasingly reflects delayed care of other diseases. In Spain, there is a huge increase in all-cause mortality, the majority reflecting a rise in COVID-19 attributable death. ¹⁵ However, Spain also experienced a rise in mortality not due to COVID-19. Could this reflect the consequence of delays in care for cancer, cardiovascular disease, and other conditions? America with 4% of the world’s population has ∼28% of all reported COVID-19 cases globally. A June 1^st New York Times article by Denise Lu summarized recent outcomes in two of the hardest hit states, New York and New Jersey. ¹⁶ While 60% of the observed excess mortality between March 15, 2020, and May 2, 2020, was due to COVID-19, 40% reflected death from other diseases, ¹⁶ notably a three-fold increase in deaths from heart disease. ¹⁶ Gogia et al. documented the magnitude of the deferred care crisis in New York City, noting an 88% drop in non-COVID-19 intensive care unit admissions from February 15, 2020, to April 15, 2020. ¹⁷ They reported a decline in admissions for myocardial infarction, stroke, and cardiac arrest. ¹⁷ These are not conditions that permit “deferred care.” Care deferred is care denied, and the unintended ethical implications of this fall in admission is likely that the “missing” patients died or suffered adverse outcomes. At Kingston Health Sciences Centre, we too saw a ∼40% dip in endovascular stroke treatment in March and April 2020. Gogia notes that even as COVID-19 admissions decline in New York City, cardiac patients are slow to return. Fear of the contagion likely drives system avoidance. In addition, widespread unemployment has further contributed to lack of hospital care access in an American healthcare system that is largely based on private insurance tied to employment. Thus, Canada cannot simply passively provide care for the 99.9%. To equitably deliver care, we will need clear policy and frequent messaging that it is safe to come to hospitals and clinics to get needed care. We must eliminate barriers to access to achieve the goal of providing equitable access to care for non-COVID-19 and COVID-19 patients. To accelerate the ramp up, the government’s policies should be guided by epidemiology and be regionally nuanced. As we begin the elective care ramp-up, Kingston Health Sciences Centre struggles with a physical plant that was not designed for management of patients with infectious diseases. The need for physical distancing limits the number of patients seen. Policies to build pandemic-proof healthcare infrastructure which support both video healthcare and physical distancing for in-person care would likely be favoured by a cost/QALY metric.

Dr. Vivek Goel has highlighted the inseparable consequences of the COVID-19-related slowing of healthcare and the economy. He states, “So often the shutdown gets framed as a debate between health and the economy, but the economy is health, too.” ¹⁸ He also notes that unemployment increases the risk of death by 1.7%. When we think of unemployment as a disease, this offers yet another reminder why we cannot sustain a singular focus on the care of patients with COVID-19. Ontarians have been directly impacted by pandemic-related shutdowns through job losses (1.1 million) or temporary layoffs or reduced hours (1.1 million). ¹⁹ Ontario unemployment rates rose to 11.3% in April 2020. A staggering 52,700 hospital procedures had been cancelled or avoided as of April 22, and an estimated 12,200 more procedures are delayed each week we continue in our current mode. ²⁰ Schneeweiss and Murtaugh summarized the financial impact of COVID-19 of non-pharmacologic interventions to contain the spread of COVID-19. ²¹ All economies, with the exception of China and Indonesia, are predicted to experience contraction of gross domestic product and a rise in unemployment. Canada’s gross domestic product is predicted to do experience a 9% fall. ²¹ Limestone Analytics estimates that COVID-19 decreased Ontario’s gross domestic product 26% in May, ²² a provincial loss exceeding $40 billion. These economic data counterbalance any complacency in our plans to reopen society.

So where does this leave us? The need to provide care for the 99.9%, while fighting COVID-19 and opening the economy, will rest on sound epidemiologic and economic data and should invoke creative, targeted measures. We will need to target solutions to long-term care facilities, focus interventions on epidemiologic hot spots, and make health policy that is nationally aware but locally informed. We should more often weigh cost/QALY when evaluating policy alternatives. Finally, until we have a vaccine or effective therapy, we will need to accept a finite risk from COVID-19 as healthcare and society ramp up. It is noteworthy that we are not dealing with the “ethics” of polio or small pox care because science has eliminated these threats. So let us consider the root causes of disease, the social determinants of health, and the value of robust medical science as we rebuild our healthcare system and reopen society. Meanwhile, we should endorse the concept that care of patients without COVID-19, the 99.9%, should neither be advantaged nor disadvantaged by healthcare policies.