The Mean Unfulfilled Lifespan (MUL): A New Indicator of a Disease Impact on the Individual Lifespan

Abstract

Declines in life expectancies provide intuitive indicators of the impact of COVID-19 on the individual lifespan. Derived under the assumption that future mortality conditions will indefinitely repeat those observed during a reference period, however, life expectancies’ intuitive interpretation becomes problematic when that reference period is only a temporary phase in the diffusion of an epidemic.

To avoid making any assumption about future mortality, I propose measuring instead the Mean Unfulfilled Lifespan (MUL), defined as the average difference between the actual and otherwise expected ages at death in an actual death cohort. For fine-grained tracking of the pandemic, I also provide an empirical shortcut to MUL estimation for small areas or short periods.

I estimate quarterly MUL values for the first half of 2020 in 144 national populations and 122 sub-national populations in Italy, Mexico, Spain and the US. Across national populations, the highest quarterly values were reached in the second quarter in Peru (3.90 years) and in Ecuador (4.59 years). Higher quarterly values still were found in New York and New Jersey, where individuals died respectively 5.41 and 5.56 years younger on average than their expected age at death.

Using a shorter, seven-day rolling window, I estimate the MUL peaked at 7.32 years in Lombardy, 8.96 years in Madrid, and 8.93 years in New York, and even reached 12.86 years for the entire month of April in Guayas (Ecuador). These results illustrate how the MUL provides an intuitive metric to track the pandemic without requiring assumptions about future mortality.

Introduction

In the past few months, the numbers of deaths from the novel coronavirus disease 2019 (COVID-19) have become part of the daily news cycle the world over. Impressive though these numbers are, they may not convey a clear sense of the scale and pace of the pandemic. By contrast, declines in life expectancies induced by COVID-19 mortality provide a simple and intuitive metric.

An aggregate indicator of period mortality conditions over the lifespan, the period life expectancy at birth (PLEB) is relatively insensitive to mortality changes at older ages. In high-income countries, where mortality at young ages is already low, recent changes in PLEB have been in the order of +.2 years annually.¹ With the notable exception of periods of armed conflict,² declines in PLEB have become rare and similarly modest. In the US, for instance, the most recent reversals in the annual PLEB gains are the .3 of a year decline during the opioid-overdose crisis, from 78.9 to 78.6 years between 2014 and 2017, and the earlier .3 of a year decline, from 75.8 to 75.5 years between 1992 1993, at the peak of the HIV epidemic.³

The impact of COVID-19 mortality on the 2020-PLEB can be expected to be substantially larger than those, in the US as well as in a number of Latin American countries. The 400,000 COVID-19 deaths in the US by December 31^st that the University of Washington’s Institute for Health Metrics and Evaluation (IHME) currently projects would translate into a 2020-PLEB reduction of nearly one-and-a-half year. Based on this set of projections, 2020-PLEB reductions would exceed two-and-a-half years in Peru and Ecuador.⁴

Moreover, estimates of PLEB reductions are sensitive to the scale of the population and to the length of the period they refer to. By averaging out COVID-19 mortality conditions in the least and most affected areas, national figures may conceal large within-country differences, especially in countries spread on large territories like the US, Brazil or Mexico, not to mention China or India. The 2020-PLEB similarly averages out mortality conditions before the first COVID-19 death and during the most severe months of the pandemic.

Estimating PLEB reductions for smaller areas and during shorter periods would thus achieve a double objective. First, tracking the pandemic at a finer-grained geographical and temporal scale should provide better insights on the pandemic than annual, national averages. Second, expected to be several times larger than these averages, estimates of PLEB reductions for the most affected areas during the most intense phase of the pandemic may receive more public attention. This is important because public awareness is critical to the participation on which mitigating policies depend.

Indeed, some estimates of PLEB reductions have reached double-digit figures (in years).^{5, 6, 7} A routinely acknowledged but easy to miss limitation of these estimates, however, is that the intuitive interpretation of the PLEB as a measure of the individual lifespan may no longer apply. The PLEB estimates the expected age at death of a newborn experiencing the mortality conditions of the reference period during her entire lifetime. The 2020-PLEB for instance assumes an annual re-occurrence of the 2020 swings in mortality induced by COVID-19 or another pandemic with a similar mortality impact. One may hope that such an annual occurrence will not become a “new normal,” but this possibility cannot be entirely ruled out on principle either. PLEB estimates for shorter periods and smaller areas similarly estimate the age at death of a newborn experiencing the conditions in that short period and small area for her entire lifetime. The highest PLEB reductions are thus estimated on the rather sinister assumption of a Groundhog-Day-like⁸ time loop repeating the worst week of the COVID-19 crisis in some of the worst affected areas, in which those born there remain as long as they live.

Mindful of the importance of a both interpretable and scalable, over space and time, measure of the impact of COVID-19 on the individual lifespan, I suggest an alternative measure to the reduction in PLEB induced by a new cause of death, the Mean Unfulfilled Lifespan (MUL). The MUL can be interpreted as the average difference between the actual age at death and the expected age at death in the absence of a specific cause among population members dying during a reference period. Making no assumption about future mortality, the MUL can be estimated for populations of any size and for periods of any length, as illustrated here with COVID-19 mortality data from 144 national populations and 122 sub-national populations in Italy, Mexico, Spain and the US, for each sex and each of the first two quarters of 2020. I also show that the MUL can be derived as the product of (1) the proportion of deaths in the population during a reference period, that are due to a specific cause and (2) the average counterfactual life expectancy in the absence of that cause among individuals who died from that cause during the reference period. In the case of COVID-19, I show that for a given population its value only changes very slowly over time, providing an easy short-cut for fine-grained tracking of the pandemic.

Conceptual Detour

Assessing PLEB reductions induced by a specific cause requires two period life tables, one representing the prevailing mortality conditions and another one representing the counterfactual mortality conditions expected in the absence of that cause. The assessment involves a relatively copious amount of life table manipulations, but decades ago Nathan Keyfitz provided most useful insights as to what these manipulations boil down to. Considering the related issue of estimating the increase in PLEB brought by the permanent elimination of a cause of death, he summarized that the increase “depends on the average time that elapses before the persons rescued will die of some other cause.”⁹ Conversely, the decrease induced by a new cause of death depends on the average time that would have elapsed before the persons who died from the new cause would have died from other causes.

This average time can be derived from the synthetic cohort approach modelled in the period life table where each death at age a from a cause C, d^C(a), reduces the number of person-years lived by the life expectancy at age a in the absence of that new cause, e_a ^o-C. This assumes that persons dying from the new cause would have had the same life expectancy in the absence of that cause as same-age persons who survived that cause. This common assumption may appear unlikely, and interactions between causes of death can be incorporated instead, but the data requirements are substantial. Under the common assumption, the difference in PLEB is thus the average over all members of the synthetic cohort, l₀ (the radix of the life table), of the difference in person-years lived by cohort member and is estimated so:

$$\Delta \mathit{\text{PLEB}}=\frac{1}{l_0}.{\int }_0^{\omega }\left(d^c(a).e_a^{o-C}\right)da$$

Keyfitz’ insight relates to the concept of “potential years of life lost”¹⁰ developed a couple of decades earlier still. In burden-of-disease assessments, it has become customary to measure the number of Years of Life Lost (YLL) as:

$$YLL={\int }_0^{\omega }\left(D^C(a).e_a^{o*}\right)da$$

where D^C (a) is the number of deaths from a certain cause C at age a observed in the population during a given period and e_a^o* is life expectancy at age a in a counterfactual period life table. Estimates of YLL due to COVID-19 are already available.^{11, 12, 13}

Three differences between the two above equations can be observed. First, the YLL uses actual numbers of deaths by age rather than life table decrements. This implies that unlike the difference in PLEB, the YLL is sensitive to the age distribution of the population, as in turns it affects the distribution of deaths by age. The standardization introduced by replacing actual deaths by life table decrements suits Keyfitz’ thought experiment comparing two states: one in which the population is indefinitely subjected to the current mortality conditions and one in which it is indefinitely subjected to mortality conditions in which a given cause of death has been permanently eliminated. As explained above, however, this approach is not well suited for tracking a fast-moving new disease such as COVID-19.

The second difference refers to the counterfactual life expectancies. In global burden-of-disease assessments, a universal life table representing optimal survival conditions is typically used. (In earlier formulation, the difference between age at death and a theoretical maximum age was used). This has the advantage of making YLL for different populations additive allowing for derivation of a global estimate of the YLL due to a cause by simple summation. However, using a universal life table may misrepresent the actual gains from averting a death in a specific population.

The last difference concerns the denominator, or lack thereof in the YLL. The use of a denominator allows for relating a total number of years in a population, measured by the YLL, to a number of persons, and thus for a more intuitive interpretation as an average difference in years lived per person. Two ratios involving the YLL can be found in the literature. First, the average YLL (AYLL) relates the YLL to all the deaths from that cause in the population during the period, D^C:

$$\mathit{\text{AYLL}}=\frac{{\int }_0^{\omega }\left(D^C(a).e_a^{o*}\right)da}{D^C}=\frac{{\int }_0^{\omega }\left(D^C(a).e_a^{o*}\right)da}{{\int }_0^{\omega }{D}^C(a)da}$$

The AYLL thus represents the average (universal) life expectancy left to population members who died from the specific cause during a given period. On the one hand, it is a very coherent measure as its denominator includes all the deaths that contribute to lost years in the numerator. On the other hand, it is only a function of the distribution of deaths by age, irrespective of the prevalence of that cause of death. The second measure is the YLL per capita.¹⁴ It depends on the prevalence of the different causes of death, but it is a less coherent measure as it includes in the denominator all the individuals in the population, including many that survived during the period and do not contribute to the YLL in the numerator.

Considering the advantages and limitations of the extant measures, I propose to add one measure of the impact of a cause of death on the individual lifespan. This measure, the Mean Unfulfilled Lifespan (MUL), can be defined as:

$$MUL=\frac{1}{D}.{\int }_0^{\omega }\left(D^C(a).e_a^{o-C}\right)da$$

where D is the total number of deaths (from all causes at all ages). Intended for situations where the underlying assumptions of PLEB might be implausible, the MUL is based, as is the YLL, on actual numbers of deaths rather than on life table decrements. Like differences in PLEB, however, it uses counterfactual life expectancies representing the mortality conditions in the population of interest, and relates the number of potential years of life lost estimated so to the total number of deaths from all causes in the population during the period. Since the radix equals the sum of all decrements at all ages in the life table, the MUL has the same structure as the difference in PLEB in the equation derived from Keyfitz’ insight and is similarly expressed as an average difference in person-years lived per person. The MUL can be understood as an average over all deaths in the population during a given period to the extent that deaths from causes other than the cause of interest could be added in the numerator but contribute nothing as long as deaths from other causes are assumed to occur at the exact age expected in the absence of the cause. Finally, the MUL differs from changes in average ages at death, which are measured by comparing deaths in different periods, and, which could actually increase if the specific cause affects people who are older on average that those dying from other causes. Interpretable as the average potential years of life lost due to a specific cause among population members dying in a certain period, the MUL complements existing indicators of the impact on the individual lifespan of a cause of death.

Empirical Shortcut

As discussed, tracking COVID-19 variations over time and across areas might be of interest, and, contrary to reductions in PLEB, calculating the MUL for small geographical areas and short periods is not conceptually problematic since it captures the actual length of lives that ended there and then, not future expectations thereof. The demand on data (including a separate counterfactual life table for each population of interest) is substantial, however, and the life table manipulations not particularly straightforward.

To simplify the estimation, the MUL can be rewritten as:

$$MUL=\frac{D^C}{D}.\frac{{\int }_0^{\omega }{D}^C(a).e_a^{o-C}da}{D^C}=\frac{D^C}{D}.\mathit{\text{AAYLL}}$$

The second term is similar to the AYLL in burden-of-disease assessments, but again based on population-specific, counterfactual life expectancies instead of universal ones, and is thus termed the Adjusted AYLL (AAYLL). In a given population, the counterfactual life expectancies are estimated from prior conditions and do not change over time. Meanwhile:

$$\frac{D^C(a)}{D^C}=\frac{M^C(a).N(a)}{M^C.N}=\frac{M^C(a)}{M^C}.\frac{N(a)}{N}$$

where M^C and M^C(a) are the all-age death rate and the death rate at age a from a specific cause, and N and N(a) are the total population size and number of individuals at age a in the population. The age distribution of deaths, D^C(a)/D^C, should also be expected to vary little within short periods because this distribution depends on the population composition, N(a)/N, and on the age pattern of cause-specific death rates, M^C(a)/M^C, both of which should vary little within short periods

This suggests that the value of AAYLL can be expected to only change slowly over time and be relatively invariant across populations with similar pre-COVID-19 life expectancies and population compositions. MUL values for a given population during a given period can then be approximated as the product of the AAYLL for the population and the all-age ratio of COVID-19 deaths to all deaths in the population during that period, D^C/D.

Materials and Methods

The equations defining the YLL and MUL look deceivingly simple. To implement them, one has to match numbers of deaths or decrements that are available or can be estimated on age intervals, with life expectancies that refer to exact ages. The changing value of life expectancy on a closed age interval can be approximated by linear interpolation, and the contribution of the interval to the YLL then equals the number of deaths or decrements in the interval times the interpolated value of life expectancy at the average age at death on the interval.¹⁵ The linear approximation is more problematic for age intervals on which mortality changes rapidly with age, and for the open-ended interval, it requires setting an arbitrary upper age limit. Unfortunately, this may concern a large share of COVID-19 deaths: in the US for instance, about 60% are above age 75 years and reported in just one closed (75 to 84 years) and one open age interval (over 85 years).

An alternative can be derived directly from estimates of all deaths by age and sex rather than with estimates of COVID-19 deaths by age and sex, distinguishing between deaths that were expected to occur in that period and those that were not (“excess” deaths), based on benchmark mortality conditions. Individuals dying in the age interval in which they were expected to die, which includes all individuals dying in the open-ended age interval, D_N+, do so a little earlier on average with the additional cause of death, as reflected by the difference between the average number of years lived after age x for individuals dying between ages x and x+n, _na_x and _na_x^−C in the period life tables with and without COVID-19,¹⁶ and for the open-ended interval, between life expectancies at age N in the period life tables with and without COVID-19 while e^o_N and e^oN^−C. Their contribution to the total decline in length of lives lived is:

$$\left(\sum _{x=0,n}^{N-n}\left({}_n{D}_x^{-C}.\left({}_n{a}_x^{-C}{-}_n{a}_x\right)\right)+\left(D_{N+}.\left(e_N^{o-C}-e_N^o\right)\right)\right)$$

where _nD_x^−C is the number of deaths between ages x and x+n that would have been expected during the same period without COVID-19. This approach has the advantage of not having to arbitrarily interpolate the value of life expectancies on the open-ended age interval.

For “excess” deaths that were not expected on the age interval, interpolating life expectancies on the age interval remains an option. A simpler alternative is to estimate the difference between the expected and actual length of life at age x, that is, as the difference between life expectancy at that age in the absence of COVID-19, e_x^o-C, and the average number of years lived after that age for those dying in the age interval, _na_x. Estimating the difference back when individuals were age x rather than at their age at death entails a small underestimation. One reason to prefer a simpler alternative that might underestimate the actual value of the difference is that using life expectancies from the life table ignores the higher prevalence of underlying long-term conditions such as obesity among individuals who die of COVID-19. The average life expectancy at the time of death for an individual dying from COVID-19 should thus be lower than for an average individual of the same age.

With the simpler approximation, the contribution to the sum of differences between the actual and expected ages at death from excess deaths between ages x and x+n in a given period can be estimated as:

$$\left({}_n{D}_x{-}_n{D}_x^{-C}\right).\left(e_x^{o-C}{-}_n{a}_x\right)$$

where _nD_x is the number of deaths from all causes between ages x and x+n in the population during the period. Adding the contributions of expected and excess deaths, and averaging across all age-groups, the MUL value can be estimated as:

$$MUL=\frac{1}{D}\left(\sum _{x=0,n}^{N-n}\left({}_n{D}_x.\left(e_x^{o-C}{-}_n{a}_x\right){-}_n{D}_x^{-C}.\left(e_x^{o-C}{-}_n{a}_x^{-C}\right)\right)+\left(D_{N+}.\left(e_N^{o-C}-e_N^o\right)\right)\right)$$

Using this approximation to estimate the impact of COVID-19 on the individual lifespan in each of the first two quarters of 2020 first requires a life table representing survival conditions in the first half of 2020 in the absence of COVID-19 whose values of e_x^o-C and _na_x^−C can be used. Combined with the number of individuals by sex and age-group, the life table values of _nm_x^−C then provide the expected numbers of deaths _nD_x^−C in the absence of COVID-19. Population data and life table functions for countries were obtained from the UN Population Division.¹⁷ Corresponding data for sub-populations in Italy, Spain, and the US were obtained from national statistical agencies.^{18, 19, 20, 21}

New life tables representing actual mortality conditions (with COVID-19) in each quarter must then be derived to calculate the corresponding values of _na_x. The construction of these life tables requires the quarterly numbers of deaths by sex and age-group. In countries where vital statistics are unavailable or incomplete, but estimates of COVID-19 deaths are available, the total number of deaths, _nD_x, can be obtained by adding these estimates (through a multi-decrement life table to adjust for competing risks of deaths) to the expected numbers of deaths in the absence of Covid-19 (_nD_x^−C). When COVID-19 estimates are not broken down by sex and age-group, an alternative is to use a reference set of age-and-sex death rates from COVID-19 from another population for which these rates are deemed reliable.²² Centers for Disease Control and Prevention (CDC) data provided the reference set of age-and-sex death rates from COVID-19.²³ Estimates of COVID-19 deaths by April 1, and July 1, 2020 were taken from the IHME.²⁴ All of these data were downloaded from institutional websites.

Results

Figure 1 shows MUL values in the first quarter (length of the dark segment of the bar) and second quarter (total length of the bar) of 2020 for both sexes in the national and sub-national populations with the largest MUL values and at least 1,000 COVID-19 deaths by July 1, 2020. As first-quarter MUL values illustrate, the individual lifespan was impacted first in parts of Italy and Spain, with limited impact elsewhere except in Ecuador. During the first quarter in Madrid, individuals died 3.16 years younger on average than their otherwise expected age at death (in the absence of COVID-19). The corresponding figure for Lombardy was 2.75 years. The average for Spain (.98 years) and Italy (.88 years) were lower than for Ecuador, however, the only country where the MUL exceeded one year during that quarter (1.29 years, Figure 1).

Figure 1:

Mean Unfulfilled Lifespan (MUL) for both sexes, by quarter and populations with 1,000 or more COVID-19 deaths by July 1^st, 2020, in years.

Between the two quarters, the MUL trended upward across Europe, including in countries where the impact had previously been limited, such as Belgium (2.81 years), and reached 3.66 years in Madrid and 3.04 years in Lombardy (Figure 1). The most notable increases, however, were in the Western Hemisphere. During the second quarter in Ecuador, individuals died 4.59 years younger on average than their otherwise expected age at death. The corresponding figure was 3.90 years in Peru. In the US, the second-quarter MUL averaged 1.76 years for the nation as a whole, but reached 5.41 in New York and 5.56 years in New Jersey (Figure 1). The above values refer to both sexes combined, and MUL values for men are even higher. During the second quarter in New Jersey, men died 6.05 years younger on average than their expected age at death before COVID-19.

To apply the suggested empirical short-cut, Appendix Table S1 provides quarterly AAYLL values using counterfactual life tables in different populations. These values range from a low of 9.62 years in Bulgaria to a high of 26.48 years in Qatar. These differences can be explained by different age compositions, with younger compositions giving more weight to remaining life expectancies at younger ages, which are obviously higher. Figure 2 shows MUL values derived from quarterly AAYLL values for a rolling seven-day period from mid-March to mid-May peaking at 7.32 years in Lombardy, 8.96 years in Madrid and 8.93 years in New York.

Figure 2:

Mean Unfulfilled Lifespan (MUL) for both sexes, by rolling seven-day periods, in years.

The same approximation but for sub-populations is illustrated by focusing on the province of Guayas in Ecuador. Data on the monthly number of deaths by province show the marked increase in March, April and May from a baseline of 1,700–2,000 per month in January, February and again in June.²⁵ In April, the number of deaths reached 12,004, of which the January-February-June average suggests only 15.5% might be estimated to be from causes other than COVID-19 (without adjustment for competing causes). No specific life table is available to estimate MUL values for the province directly, but based on the second-quarter AAYLL derived for Ecuador, individuals appeared to have died 12.86 years younger, on average, that their expected age at death in April in the province of Guayas. This probably represents the largest impact of COVID-19 on the individual lifespan to date.

Discussion

The MUL is proposed as an alternative to induced changes in PLEB to assess the impact of a cause of death of the individual lifespan in situations where the assumptions underlying life table constructions are implausible. Complementing existing measures such as the AYLL and the YLL per capita, the MUL is similarly based on actual number of deaths from a specific cause, but by averaging an estimate of potential years of life lost over the total number of deaths in the period, the MUL is structured like a difference in PLEB. To calculate the MUL, it is suggested to work with the number of deaths from all causes, attributing a different average number of potential years of life lost to deaths that would have been expected in the absence of a cause and to those that are “excess” deaths. This approach also reinforces the interpretation of the MUL as an average number of years of life lost across all deaths in a given period.

To illustrate its use with COVID-19 mortality during the first half of 2020, MUL values were estimated for a total of 266 populations for each sex and each quarter. To estimate this large number of values, the simplifying assumption of a common age-and-sex pattern of COVID-19 mortality was used (i.e., the same age-and-sex-specific death rates relative to all-age, both-sex death rate in all populations). The distribution of COVID-19 deaths by sex and age-group derived so can obviously be replaced with the actual distribution in populations for which that distribution is available.

As with YLL-based measures, an approximation is also required to estimate the difference between the actual and expected ages at death for individuals dying on an age interval (ages x to x+n). Using the average age at death on that age interval and life expectancy prior to COVID-19 at beginning of that age interval (exact age x), this difference was estimated here when individuals were at exact age x rather than at their age at death. This entails some underestimation of the difference on each age interval, which was preferred to compensate for a typical, but rarely entirely plausible assumption common to differences in PLEB estimates and YLL-based measures. This assumption is that in the absence of a certain cause of death the persons who in fact died from that cause would have had the same remaining life expectancy as any same-age person, when in the case of COVID-19 in particular, a shorter expectancy would be likely considering the higher proportion of underlying long-term conditions (e.g., obesity) among COVID-19 victims. In the United Kingdom, one study found that the estimated AYLL for COVID-19 was reduced from 13 to 12 years (average for both sex) when controlling for these conditions.²⁶ These AYLL compare to AAYLL values for the United Kingdom estimated here to be 11.85 years during the first quarter and 11.94 years during the second one (Appendix Table S1). Our AAYLL values for the USA, 12.64 years in the first quarter and 12.71 years in the second, are also consistent with an earlier estimate that the country would lose 12.3 million years of remaining life if the COVID-19 death toll was to reach one million.²⁷

With the AAYLL values provided in Appendix Table S1, MUL values can easily be approximated as the product of the corresponding AAYLL value and the ratio of COVID-19 to total deaths in the population during the period. The approximation can be used for short-duration periods as long as AAYLL can be assumed to be almost constant over time and AAYLL values estimated in each of the first two quarters were found indeed to have changed little from one quarter to the next. This approximation can also be used for sub-populations for which the necessary data are only provided of the entire population, but only when the age compositions of the sub-populations can be held as relatively similar.

This relates more generally to the fact that, like YLL-based measures, the MUL is not sex- or age-standardized and MUL comparisons across populations that differ markedly in age composition will be biased. On the one hand, all else equal, a younger population composition yields a younger distribution of deaths and a higher AAYLL value. On the other hand, known variations of COVID-19 mortality with age²⁸ suggest that an older population distribution contributes to a higher proportion of COVID-19 deaths relative to all deaths. Only by accident would the two opposite age-composition effects cancel each other and in general the MUL is not a standardized measure—a clear disadvantage compared to differences in PLEB. The PLEB is the inverse of a “stationary” death rate: a weighted average of the period age-specific death rates with weights derived from these death rates through life table construction. Using these internally-derived weights rather than an external, standard age distribution,²⁹ the stationary death rate still achieves the main goal of age-standardization, namely, to yield a mortality measure independent of the actual age composition of the population. However, this internal derivation assumes current mortality conditions will become permanent.

When this assumption is not tenable, the MUL provides an unstandardized alternative to differences in PLEB, that is similarly structured as an average difference in length of lives lived per person. Its interpretation pertains to an actual death cohort, that is, population members who died during a certain period rather than to a synthetic cohort as represented in the life table. The MUL indicates how much younger than expected, based on mortality conditions before the onset of a new cause of death, members of this death cohort died on average, and its measurement requires no assumption about future mortality trends.

Significance Statement.

To convey the significance of COVID-19 in a relatable metric is important because public awareness is critical to the participation on which mitigating policies depend. Mortality indicators are among the most salient measures of the impact of COVID-19. While demographers favor age-standardized death rates to track the pandemic, those are expressed in unintuitive metrics: deaths per 1,000 or fraction thereof. Declines in life expectancies are more intuitive indicators, but, derived under an assumption of unchanged future mortality, they are unsuitable for fine-grained tracking of a fast-moving epidemic. To avoid making any assumption about future mortality, I introduce a Mean Unfulfilled Lifespan (MUL), defined as the average difference between the actual and otherwise expected ages at death in an actual death cohort.