FormalPara Key Summary Points

The term long COVID (LC) captures the wide-ranging long-term effects of severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) infections but lacks specificity for precise research, diagnostics and treatment.

Refinement into three subgroups—ongoing symptomatic COVID-19, induced/exacerbated diseases and post-acute COVID condition—facilitates better clinical management and precise epidemiological research.

Post-acute COVID condition is an umbrella term. Individual post-acute COVID condition may include post-exertional malaise (PEM), dysautonomia, immune system dysregulation and/or cognitive dysfunctions, among others, which should be diagnosed by the physician and stratified in research.

Recognizing and addressing PEM within the post-acute COVID condition is essential for effective treatment and preventing chronification.

Commentary

The term long COVID (LC) is very well suited to describe the overall disease burden that can result from infection with severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) in an individual [1]; the personal disease burden, tremendously reduced quality of life and socioeconomic implications at the population level show the overall impact of the condition and importance of meaningful preventive measures. However, the term is not well suited to providing a tailored approach to epidemiological research, initial diagnostics and therapy because of the need for differentiation regarding the various symptoms and syndromes subsumed under the term LC to be able to offer physicians precisely targeted and effective solutions in the daily practice [2].

For example, there are always misunderstandings when interpreting frequency data on LC after a SARS-Cov-2 infection. Even without going into limitations of the different methodologies in more detail, there is a big difference between asking the population whether they experienced symptoms lasting > 4 weeks after a SARS-CoV-2 infection and using health record data as the basis for the calculations. For the population survey, all long-term effects experienced will be reported, whether due to a more severe course of COVID-19 or newly acquired symptoms, syndromes, or illnesses as a post-acute result of the infection. When using health record data, only the related diagnoses found and documented by physicians and then defined by researchers are used; this means, depending on the definition, only a part of all long-term effects is recognized. Here, the misunderstandings start because all of this is usually called LC.

Therefore, the National Institute for Health and Care Excellence (NICE) rapid guideline has already attempted an initial distinction of the term LC by delineating “ongoing symptomatic COVID-19” from the “post-COVID condition,” with the criterion that it cannot be explained by alternative diagnoses like the WHO did and additionally introduced ICD-10 code U09.9 for the post-COVID condition [3, 4]. However, according to this NICE definition, ongoing symptomatic COVID-19 ends after 12 weeks, which often does not correspond to the observable reality of the disease like for formerly critically ill patients as well as conditions like severe myocarditis. How should we proceed if symptoms of diseases were induced, triggered or exacerbated by SARS-CoV-2 but themselves fall under the alternative diagnosis of, e.g., autoimmune, cardiovascular or neurodegenerative disease? An extensive recent report by the National Academies of Sciences, Engineering and Medicine comprehensively outlines the various health impacts associated with LC [5]. The findings indicate that LC is a multifaceted chronic condition, manifesting in > 200 different health effects across multiple organ systems [5]. These effects encompass both new onset symptoms and the exacerbation of pre-existing conditions [5]. A recent publication by Ewing et al. (2024) additionally described LC as a disease with a spectrum of pathology [6]. They recommended even broadening the definition of long-term effects not only to include diseases and organ damage caused by SARS-CoV-2 but also (currently still) asymptomatic organ damage, which can lead to various health impacts such as increased risk for heart attacks and strokes in the future [6]. Xie et al. also calculated the risk and burden of long-term effects of a SARS-CoV-2 infection by including many organ systems and related diseases, naming it post-acute sequelae of SARS-CoV-2 infection (PASC) [7]. This term is also widely used in scientific literature and is probably, together with the more colloquial term LC, the adequate term for the whole disease spectrum from acute to post-acute when it comes to pathomechanism research and showing the full impact of damage.

However, this is generally not what doctors mean by LC when they see patients, since probably few doctors would describe, document and code, for example, a heart attack or autoimmune disease induced by a previous SARS-CoV-2 infection as LC, PASC, or post-COVID condition. It has not been called post-bacteremia syndrome in the past if a patient developed a stroke during bacteremia. What doctors are most likely to diagnose as LC is what NICE [3] has summarized as ongoing symptomatic COVID-19 or post-COVID condition as well as what the ESCMD rapid guideline describes [8]. Nevertheless, tailored terms are additionally needed for the other long-term effects mentioned above.

Therefore, we propose to use the term LC or PASC only if aiming at including all long-term effects, and the terms for the three subgroups are described in Fig. 1 for more targeted sampling or stratification approaches, which do not include all long-term effects. Additionally, we propose to consider the three subgroups for a more targeted diagnostic strategy if a patient with long-COVID symptoms comes into the doctor's office. We further differentiated the definition of the third subgroup in a second step.

  1. 1.

    Group: Ongoing symptomatic COVID-19: Prolonged courses of severe COVID-19, e.g., acute respiratory distress syndrome (ARDS), severe pneumonia, myocarditis, multisystem inflammatory syndrome (MIS), symptoms due to organ damage during severe COVID-19 (e.g., brain, heart, kidney, lung damage) or therapy (e.g., ventilation, post-ICU syndrome) with symptoms that persist up to 12 weeks. After 12 weeks, we suggest to name it “(fibrosis as) ongoing sequelae of COVID-19,” for example.

  2. 2.

    Group: SARS-CoV-2 induced, triggered or exacerbated diseases. Symptoms for > 4 weeks due to worsening of an existing pre/chronic/residual disease due to SARS-CoV2 infection and/or post-acute new onset of diseases associated with SARS-CoV2 infection (e.g., thrombosis, embolism, stroke, cardiovascular disease, asthma, atypical pulmonary fibrosis, autoimmune disease, diabetes, dementing syndromes) and/or related increased risks. An example would be “SARS-CoV-2 triggered asthma bronchiale” or “SARS-CoV-2 exacerbated lupus erythematosus.”

  3. 3.

    Group: Post-acute COVID condition (cannot be explained by another diagnosis like the ones above) instead of post-COVID condition to make it clearer that this is an expression of a post-acute infection syndrome caused by SARS-CoV-2 with symptoms that persist for > 3 months but could be present even before.

Fig. 1
figure 1

Description of the terms for the three proposed long-COVID subgroups over time

Full size image

Since LC/ PASC is a spectrum, the three subgroups can also occur in different combinations and influence each other.

Most important in the context of the sub-group building is that for the first two subgroups in contrast to subgroup 3, there are already known, clearly graspable diagnostic markers and, for the most part, guideline-compliant treatment options as shown for example for neurological symptoms by Frontera et al. [9]. This means that symptoms due to a possible affiliation to the first two subgroups must be excluded using excellent but common differential diagnostics to then take an equally careful diagnostic approach for subgroup 3.

To differentiate the first subgroup would additionally make sense in terms of known risk factors for LC, which overlap largely with risk factors for severe acute disease progression [10]. Prolonged courses and immediate organ damage resulting from the acute disease occur mainly in severe acute infections. The first subgroup of LC might currently be most amenable to prevention if severe courses of acute COVID-19 can be limited or prevented as much as possible. This is also precisely the subgroup for which vaccination, medication and the reduction of other factors responsible for the risk of a severe COVID-19 course including MIS have already been shown to reduce the risk of LC [11]. These considerations are in line with previous comments about the differentiation of LC symptoms: “Long haulers who survived severe acute SARS-CoV-2 infection are most likely to be men older than 50 years with lingering tissue damage and scarring (subgroup 1; added by authors). People with long COVID after a less severe infection are most likely to be younger women (aged 36–50 years) whose acute infection has triggered adverse physiological responses (subgroup 3; added by authors)" [2]. Indeed, it was shown that the decline of the overall LC frequencies after infection between 2020 and 2022 was the result of two key drivers: availability of vaccines and changes in the characteristics of the virus—which made the virus less prone to causing severe acute infections and related long-term effects [7].

Subgroup 2 includes LC symptoms caused by the diseases for which a significant increase in risk could be shown because of SARS-CoV-2 infection. Importantly, first hints regarding the cumulative risk of re-infections in particular in subgroup 2 are reported in the literature [12, 13]. Therefore, in particular the number of individuals with a new onset or worsening of disease might increase over time as new patients with LC symptoms occur because of ever-new infections with novel variants cycling through the population. A challenge for subgroup 2 is that the new disease onset might also be after 12 weeks, and links with the causative SARS-CoV-2 infection might be increasingly difficult to identify and document as sequelae of the SARS-CoV-2 infection, in particular if clinicians are not used to link responsible pathogens to long-term effects. However, a recent study showed that even 3 years after the initial infection, new diseases, syndromes and symptoms can occur [14]. Therefore, it might be useful to implement it in the documentation of diseases linked to certain pathogens.

For subgroup 3, if taking the definition of “cannot be explained by another diagnosis” seriously, it becomes increasingly clear that standard diagnostics and treatment options are usually not fit for purpose and that this is the group with the most similarities to previously known post-acute infection conditions. It was previously stated that “…post-acute infection syndrome with clear physiological dysfunction…is often not consistently apparent using standard medical diagnostic tests. This discrepancy highlights the need for a new generation of more sensitive testing procedures for people with post-acute infectious syndrome" [2]. In line with this, we suggest using post-acute-COVID condition instead of just post-COVID condition for the post-acute infection syndrome (PAIS) triggered by SARS-CoV-2 [15]. Moreover, this is the subgroup of long-COVID which occurs often even after a mild SARS-CoV-2 infection [2].

PAIS are not new [15]. The term PAIS, however, is an umbrella term for dysfunctions and damage that can occur after the acute phase of illness with certain pathogens. Individual post-acute infection symptoms and syndromes already identified are for example [15]: post-exertional malaise (PEM), dysautonomia (DA), e.g., postural tachycardia syndrome (PoTS), dysregulation of the immune system, e.g., mast cell overactivation (MCAS)-like syndromes, and cognitive dysfunctions. This is exactly where the stratification of this group should start when it comes again to research, diagnosis and therapy. Most importantly, we suggest always differentiating between patients with post-acute COVID condition with and without the existence of PEM. The presence of PEM makes an important difference when it comes to treating patients [16]. Unlike all other patient groups, patients with PEM must first have their condition stabilized by pacing [1, 17]. Pacing must be considered in all further forms of treatment and always come first. This is a major challenge since graduate activation therapy is a successful form of therapy, e.g., for PoTS when occuring alone, but must not be treated according to standard autonomic dysfunction guidelines in patients with PEM being also present [18]. Patients with post-acute COVID condition with PEM are, therefore, particularly vulnerable and at high risk of suffering a chronic course as is seen in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) type of post-acute COVID condition with its hallmark symptom PEM [1, 17].

Figure 2 illustrates the mainly assumed mechanisms and related symptoms that should be considered concerning more sensitive new and modern diagnostic techniques regarding post-acute COVID condition. At the center of consideration, special relevance has to be given to the occurrence of PEM [16,17,18,19,20,21].

Fig. 2
figure 2

Mainly assumed mechanisms and related symptoms of post-acute COVID condition that should be considered concerning more sensitive new and modern diagnostic techniques

Full size image

Moreover, as stated before, the current definition and linked International Classification of Diseases, Tenth Revision (ICD-10) code raise the question of correct documentation and coding options for long COVID symptoms as shown in the latest data from Statistics Canada, in which only 5.7% of LC sufferers received a corresponding diagnosis [22]. How are the symptoms of subgroups 1 and 2 documented and coded if the existing ICD-10 code U09.9 is only to be used for the post-COVID condition, although subgroups 1 and 2 are also sequelae of the SARS-CoV-2 infection?

We suggest coding the respective disease and/or symptom as the principal diagnosis for the first two subgroups and U09.9 as the secondary diagnosis, in contrast to coding U09.9 as the principal diagnosis for subgroup 3, the post-acute COVID group and the corresponding symptoms and syndromes subsumed under the term post-acute COVID, such as PoTS with G90.80 or ME/CFS with G93.3, as the secondary diagnosis. For example, registry studies and studies that use medical records as the basis for their data could, then, clearly show which damage is caused by SARS-CoV-2 as a whole and which is caused by subgroup 3 alone, the post-acute infection syndrome group, provided that the differential diagnosis by the doctors carrying out the diagnostics and their documentation is correct and carefully carried out. This would be feasible for other pathogens, too, where also a long-term overview of the damages is needed.

In summary, we recommend utilizing the term "LC" or "PASC" when referring to the comprehensive spectrum of long-term effects. For more precise sampling or stratification that does not encompass all long-term effects, the terms associated with the three subgroups should be employed. On the one hand, doctors should clearly differentiate the diagnostic pathways for patients to be able to assign risk factors more clearly. On the other hand, this subdivision is important to have a basis for uniform clinical coding for epidemiological research, for example. In addition, the third subgroup, the post-acute infectious syndrome post-acute COVID condition, must also be further stratified since it is essential for diagnostic and therapeutic developments. It is especially important to consider the presence of PEM.

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.