The COVID-19 epidemic burst in Bergamo region for the first time after China, in early 2020, provoking an unprecedented crisis of the hospital system and the society as a whole, intended to last for many months to follow(25, 26). Our intention was to study if pre-existing socioeconomical disadvantage could shape the individual recovery.
To do so, we employed a large database from patients followed-up after receiving hospital care (not forcefully admitted, even though in the period under study our health system was in such a complete distress, for shortage of hospital resources, that a similar distinction would lose significance).
The completeness and quality of recovery, after at least 12 weeks from onset, was in-person investigated by Infectious Diseases or Internal Medicine specialists (for symptoms assessment), trained Psychologists (by means of semiquantitative scales for PTSD and HRQoL), and Physical Therapists (for Barhel scale and BFI); DLCO was measured by a Respiratory Medicine specialist.
We recorded a high prevalence of symptoms and pathologic results in BFI, Barthel’s scale, IES-R, SF-36, and DLCO measurement: those findings are perfectly in line with others’ reports)(27–29).
Among these outcomes, only HRQoL resulted significantly associated with social disadvantage, specifically for its items addressing the physical dimension. This association was independent from sex, age, BMI, number of comorbidities and time to follow-up.
Our findings support the idea that social disadvantage is a strong determinant of the recovery process, after acute infection by SARS-CoV-2. This could be shaped by the reduced access to healthcare, or by its poorer quality. In addition, pre-existing or ongoing behavioural and dietary factors could be responsible. For employed individuals, the impossibility of staying off from work for long periods (or the higher physical efforts required by poorer working conditions) could be in cause; similarly, for retired patients, a reduced access to home care and assistance could mark the difference. Research should be pursued in this direction.
In contrast with other studies(3–7), we could not find any association between social disadvantage and acute-phase COVID-19 severity, and this is reasonably due to the choice of excluding those patients, who experienced the most relevant acute complications.
A rich literature has flourished about PASC, but it would be inaccurate to label as “PASC” the clinical condition that we observed at follow-up, because, at the time when we started our intervention, no formal definition of PASC had already been established: for this reason, we adopted the composite endpoint of “Physical symptoms” (which, importantly, does not consider minor cognitive deficits). Anyway, even if our results are note directly transferrable to PASC conditions, it is notable that very few authors have studied how pre-existing SES is associated to PASC development, while in general it is recognized that PASC has a relevant impact on social functioning(30), working capacity(10, 31) and household finances:
-
Yoo and collaborators(32) investigated the effects by SES on PASC, but could not find any. In their study, though, PASC was defined through an ad hoc questionnaire, incorporating questions from SF-36, but not specifically targeting HRQoL. Unlike them, we examined symptoms, psychologic scales and HRQoL (SF-36), each independently.
-
More recently, authors from Sao Paulo, Brazil(33), found an association between symptoms at follow-up (not fulfilling the formal definition of PASC) and socio-economic deprivation.
-
A multicentre study on the influence of SES upon functional recovery, after ICU admission for COVID-19-related ARDS, is ongoing in France(34).
-
According to a big-data analysis by the UK Office for National Statistics, a higher prevalence of “long-COVID” (another term referring to PASC) is found in the most deprived areas of the country (as by Index of Multiple Deprivation)(35).
Our analysis has many strengths. First of all, the population studied is unique: for the “catastrophic” nature of the events experienced, at the very beginning of the COVID pandemic and at a time when all the social inequalities produced by the pandemic had not yet firmly established. In addition, the study was entirely done in a pre-vaccine “era”, when the circulating viral variant was still the original one. The sample considered is large and well balanced among social classes, actively and systematically recruited, and directly interviewed by multidisciplinary staff, and the model we adopted, to estimate SES, is strong and validated by the main Italian Institute for demographic studies (ISTAT). Finally, the semi-quantitative measurement of HRQoL, here obtained by specifically trained professionals, accounts for a more reproducible and accurate(36) assessment than the mere symptoms list, as recognized also in other “chronic fatigue conditions”(37, 38).
We acknowledge the following weaknesses: a scarce representativeness of the whole population affected by SARS-CoV-2 (not admitted to hospital, nor consulting the Emergencies: our study is mainly focused on post-hospitalized patients); a significant attrition cascade; the unavailability of pre-COVID results for the scales adopted, and for DLCO.
Lacking a non-COVID control group, we cannot exclude that HRQoL reduction in lower classes depends on disadvantage itself, independently of the recovery from COVID. In fact, associations between HRQoL and socioeconomic disadvantage are well established, in other research settings, especially in response to acute illness (like falls in the elderly(39), or ischemic cardiac disease(40)). However, were the observed HRQoL reduction depending exclusively on socioeconomic disadvantage, one would expect it to act also on the psychological outcomes (i.e. on IES-R and on SF-36 items exploring the psychological domains), which is not apparent from our results.
A long time has passed now, since the hard times of the first COVID-19 waves all around the world, and the clinical characteristics of the disease, together with the reduced severity observed in the immunized hosts, have radically improved also the recovery process, making our results poorly transferrable to the current scenario.
Nonetheless, the impact of socio-economic inequalities upon such a traumatic occurrence - as the first wave of COVID-19 has been everywhere – deserved a special attention by researchers, because similar events are far from impossible to happen again.