Using 'Pacing' to Manage Chronic Fatigue Syndrome: Insights for Long COVID

Post-viral illness occurs when individuals experience an extended period of feeling unwell after a viral infection [1–6]. While post-viral illness is generally a non-specific condition with a constellation of symptoms that may be experienced, fatigue is amongst the most commonly reported [7–9]. For example, our recent systematic review found there was up to 94% prevalence of fatigue in people following acute COVID-19 infection [3]. The increasing prevalence of long COVID has generated renewed interest in symptomology and time-course of post-viral fatigue, with PubMed reporting 72 articles related to “post-viral fatigue” between 2020 and 2022, but less than five for every year since 1990.

As the coronavirus pandemic developed, it became clear that a significant proportion of the population experienced symptoms which persisted beyond the initial viral infection, meeting the definition of a post-viral illness. Current estimates suggest one in eight people develop long COVID [10] and its symptomatology has repeatedly been suggested to overlap with clinical demonstrations of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). In a study by Wong and Weitzer [11], long COVID symptoms from 21 studies were compared to a list of ME/CFS symptoms. Of the 29 known ME/CFS symptoms the authors reported that 25 (86%) were reported in at least one long COVID study suggesting significant similarities. Sukocheva et al. [12] reported that long COVID included changes in immune, cardiovascular, metabolic, gastrointestinal, nervous and autonomic systems. When observed from a pathological stance, this list of symptoms is shared with, or is similar to, the symptoms patients with ME/CFS describe [13]. In fact, a recent article reported 43% of people with long COVID are diagnosed with ME/CFS [13], evidencing the analogous symptom loads.

A striking commonality between long COVID and similar conditions such as ME/CFS is the worsening of symptoms including fatigue, pain, cognitive difficulties, sore throat, and/or swollen lymph nodes following exertion. Termed post exertional malaise (PEM) [14–17], lasting from hours to several days, it is arguably one of the most debilitating side effects experienced by those with ME/CFS [16–18]. PEM is associated with considerably reduced quality of life amongst those with ME/CFS, with reduced ability to perform activities of daily living, leading to restraints on social and family life, mental health comorbidities such as depression and anxiety, and devastating employment and financial consequences [19–22]. At present, there is no cure or pharmacological treatments for PEM, and therefore, effective symptom management strategies are required. This may be in part because the triggers of PEM are poorly understood, and there is little evidence for what causes PEM, beyond anecdotal evidence. The most common approach to manage PEM is to incorporate activity pacing into the day-to-day lives of those with ME/CFS with the intention of reducing the frequency of severity of bouts of PEM [23]. Pacing is defined as an approach where patients are encouraged to be as active as possible within the limits imposed by the illness [23–25]. In practice, pacing requires individuals to determine a level at which they can function, but which does not lead to a marked increase in fatigue and other symptoms [26, 27].

Although long COVID is a new condition [3, 14], the available evidence suggests substantial overlap with the symptoms of conditions such as ME/CFS and it is therefore pragmatic to consider the utility of management strategies (such as pacing) used in ME/CFS for people with long COVID. In fact, a recent Delphi study recommended that management of long COVID should incorporate careful pacing to avoid PEM relapse [28]. This position was enforced by a multidisciplinary consensus statement considering treatment of fatigue in long COVID, recommending energy conservation strategies (including pacing) for people with long COVID [29]. Given the estimated > 2 million individuals who have experienced long COVID in the UK alone [30–32], there is an urgent need for evidence-based public health strategies. In this context, it seems pragmatic to borrow from the ME/CFS literature.

From a historical perspective, the 2007 NICE guidelines for people with ME/CFS advised both cognitive behavioural therapy (CBT) and graded exercise therapy (GET) should be offered to people with ME/CFS [33]. As of the 2021 update, NICE guidelines for people with ME/CFS do not advise CBT or GET, and the only recommended management strategy is pacing [34]. In the years between changes to these guidelines, the landmark PACE trial [35] was published in 2011. This large, randomised control trial (RCT; n = 639) compared pacing with CBT and reported GET and CBT were more effective than pacing for improving symptoms. Yet, this study has come under considerable criticism from patient groups and clinicians alike [36–39]. This may partly explain why NICE do not advise CBT or GET as of 2021, and only recommend pacing for symptom management people with ME/CFS [34]. There has been some controversy over best treatment for people with ME/CFS in the literature and support groups, potentially amplified by the ambiguity of evidence for pacing efficacy and how pacing should be implemented. As such, before pacing can be advised for people with long COVID, it is imperative previous literature concerning pacing is systematically reviewed. This is because a consensus is needed within the literature for implementing pacing so practitioners treating people with ME/CFS or long COVID can do so effectively. A lack of agreement in pacing implementation is a barrier to adoption for both practitioners and patients. Despite several systematic reviews concerning pharmacological interventions or cognitive behavioural therapy in people with ME/CFS [36, 40, 41], to date, there are no systematic reviews concerning pacing.

Despite the widespread use of pacing, the literature base is limited and includes clinical commentaries, case studies, case series, and few randomised control trials. Consequently, while a comprehensive review of the effects of pacing in ME/CFS is an essential tool to guide symptom management advice, the available literature means that effective pooling of data is not feasible [42] and therefore, a traditional systematic review and meta-analysis, with a tightly focussed research question would be premature [43]. Consequently, we elected to undertake a scoping review. This approach retains the systematic approach to literature searching but aims to map out the current state of the research [43]. Using the framework of Arksey and O'Malley [44], a scoping review aims to use a broad set of search terms and include a wide range of study designs and methods (in contrast to a systematic review [44]). This approach, has the benefit of clarifying key concepts, surveying current data collection approaches, and identifying critical knowledge gaps.

We aimed to provide an overview of existing literature concerning pacing in ME/CFS. Our three specific objectives of this scoping review were to (1) conduct a systematic search of the published literature concerning ME/CFS and pacing, (2) map characteristics and methodologies used, and (3) provide recommendations for the advancement of the research area.

The review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews (PRISMA-ScR) guidelines [45] and the five-stage framework outlined in Arksey and O’Malley [44]. Registration is not recommended for scoping reviews.

Studies that met the following criteria were included in this review: (1) published as a full-text manuscript; (2) not a review; (3) participants with ME/CFS; (4) studies employed a pacing intervention or retrospective analysis of pacing or a case study of pacing. Studies utilising sub-analysis of the pacing, graded activity, and cognitive behaviour therapy: a randomised evaluation (PACE) trial were included as these have different outcome measures and, as this is not a meta-analysis, this will not influence effect size estimates. Additionally, due to the paucity of evidence, grey literature has also been included in this review.

The search strategy consisted of a combination of free-text and MeSH terms relating to ME/CFS and pacing, which were developed through an examination of published original literature and review articles. Example search terms for PubMed included: ‘ME/CFS’ OR ‘ME’ OR ‘CFS’ OR ‘chronic fatigue syndrome’ OR ‘PEM’ OR ‘post exertional malaise’ OR ‘pene’ OR ‘post-exertion neurogenic exhaust’ AND ‘pacing’ OR ‘adaptive pacing’. The search was performed within title/abstract. Full search terms can be found in Additional file. 1

Six electronic databases [PubMed, Scholar, ScienceDirect, Scopus, Web of Science, and the Cochrane Central Register of Controlled Trials (CENTRAL)] were searched to identify original research articles published from the earliest available date up until 02/02/2022. Additional records were identified through reference lists of included studies. ‘Grey literature’ repositories including MEPedia, Action for ME, and ME Action were also searched with the same terms.

Once each database search was completed and manuscripts were sourced, all studies were downloaded into a single reference list (Zotero, version 6.0.23) and duplicates were removed. Titles and abstracts were screened for eligibility by two reviewers independently and discrepancies were resolved through discussion between reviewers. Subsequently, full text papers of potentially relevant studies were retrieved and assessed for eligibility by the same two reviewers independently. Any uncertainty by reviewers was discussed in consensus meetings and resolved by agreement. Data extracted from each study included sample size, participant characteristics, study design, trial registration details, study location, pacing description (type), intervention duration, intervention adherence, outcome variables, and main outcome data. Descriptions were extracted with as much detail as was provided by the authors. Study quality was assessed using the Physiotherapy Evidence Database (PEDro) scale [46, 47].

The study sponsors had no role in study design, data collection, analysis, or interpretation, nor writing the report, nor submitting the paper for publication.

After the initial database search, 281 records were identified (see Fig. 1). Once duplicates were removed, 177 titles and abstracts were screened for inclusion resulting in 22 studies being retrieved as full text and assessed for eligibility. Of those, five were excluded, and 17 articles remained and were used in the final qualitative synthesis.

Study characteristics are summarised in Table 1. Of the 17 studies included, three were randomised control trials (RCTs [35, 48, 49]); one was an uncontrolled trial [50]; one was a case series [51]; one was a retrospective observational study [52], two were prospective observational studies [53, 54]; four were cross-sectional observational studies [25, 55, 56]; and five were cross-sectional analytical studies [57–61] including sub-analysis of the PACE trial [35, 56, 59, 61]. Seven of the studies were registered trials [35, 48–50, 56–58]. Diagnostic criteria for ME/CFS are summarised in Table 2.

The majority of studies (12/17) [25, 48, 50–56, 58, 60, 63] highlighted improvements in at least one outcome following pacing (Fig. 2). When the effect of pacing was assessed by ME symptomology and general health outcomes, studies reported pacing to be beneficial [25, 50, 51, 53–56, 58]. It is worth noting however that pacing reportedly worsened ME symptoms in 14% of survey respondents, whilst improving symptoms in 44% of respondents [52]. Most studies using fatigue as an outcome measure reported pacing to be efficacious (7/10) [50, 51, 53, 54, 56, 60, 63]. However, one study reported no change in fatigue with a pacing intervention (1/10) [35], and 2/10 studies [53, 63] reported a worsening of fatigue with pacing. Physical function was used to determine the efficacy of pacing in 11 studies [35, 48, 50, 51, 53, 54, 56, 58–60, 63]. Of these, the majority found pacing improved physical functioning (8/10) [48, 50, 51, 53, 54, 56, 58, 60], with 1/10 [35] studies reporting no change in physical functioning, and 1/10 [59] reporting a worsening of physical functioning from pre- to post-pacing. Of the seven studies [35, 49–51, 53, 54, 60] which used pain to assess pacing efficacy, 4/7 [50, 51, 53, 60] reported improvements in pain and 3/7 [35, 51, 53] reported no change in pain scores with pacing. All studies reporting quality of life (1/1) [53], self-efficacy (3/3) [50, 53, 59], sleep (2/2) [53, 59], and depression and anxiety (4/4) [50, 53, 59, 63], found pacing to be efficacious for ME/CFS participants.

The majority of studies (10/17) [25, 50, 52–54, 58–61, 63] did not report age of the participants. For those which did report age, this ranged from 32 ± 14 to 43 ± 13 years. Where studies reported sex (11/17) [35, 48–51, 54–58, 60], this was predominantly female, ranging from 75 to 100% female. Only six studies [35, 54, 56–58, 60] reported ethnicity, with cohorts predominantly Caucasian (94–98%). Time since diagnosis was mostly unreported (12/17) [25, 48–50, 52–54, 58–61, 63] but ranged from 32 to 96 months, with a cross-sectional survey reporting 2% of the participants were diagnosed 1–2 years previously; 6% 3–4 years since diagnosis; 13% 3–4 years since diagnosis; 12% 5–6 years since diagnosis; 20% 7–10 years since diagnosis; 29% 11–21 years since diagnosis; 13% 21–30 years since diagnosis; and 5% > 30 years since diagnosis. Of the studies which reported comorbidities of the participants (6/17) [25, 35, 50, 56, 57, 63], the comorbidities were chronic pain, depressive disorder, psychiatric disorder.

Of the 17 studies, 14 were from Europe [25, 35, 48–52, 55–59, 61, 63], and three from North America [53, 54, 60]. Of the 14 studies[25, 35, 48–52, 55–59, 61, 63] from Europe, ten [25, 35, 50, 52, 55–59, 61, 63] were conducted in the United Kingdom, three in Belgium [48, 49, 51], and one was a multicentred study between the United Kingdom and Norway [58].

Of the 17 studies, three [53, 54, 60] used announcements in a newspaper and physician referrals to recruit participants, two [50, 63] recruited patients referred by a consultant from a National Health Service (NHS) Trust following a pain diagnosis, two [52, 55] concerned online platforms on the web, two [59, 61] recruited from secondary care clinics, and two used the PACE trial databases [56, 57]. Moreover, one study recruited from the hospital [58], one from physiotherapist referrals [25], two from specialist clinic centres [35, 64], one from waiting list of rehabilitation centre [48], and one from medical files [49].

Ten studies were carried out in hospital and clinic setting [25, 35, 48–51, 58, 59, 61, 63]. Two studies were performed on online platforms [52, 55]. Three studies did not report study setting [53, 54, 60]. Two studies generated output from PACE trial databases [56, 57]

All five intervention studies reported adherence rates (which they defined as number of sessions attended), which ranged from 4–44% (4% [49], 8% [35], 25% [48], 29% [51], and 44% [50]). One study reported the median number of rehabilitation programme sessions attended was five out of six possible sessions, with 58.9% [50] participants attending ≥ 5 sessions; 83.2% participants attending at least one educational session on activity pacing and 56.1% attending both activity pacing sessions.

Most studies (12/17) highlighted improvements in at least one outcome measure following pacing. Pacing was self-reported to be the most efficacious, safe, acceptable, and preferred form of activity management for people with ME/CFS [55]. Pacing was reported to improve symptoms and improve general health outcomes [25, 50, 52, 58, 63], fatigue and PEM [48, 50, 51, 53–56, 60, 63], physical functioning [48, 50, 51, 53, 56, 58, 60, 63], pain [25, 50, 55, 63], quality of life [50], self-efficacy [50, 53], sleep [53, 55], and depression and anxiety [50, 53, 63]. These positive findings provide hope for those with ME/CFS, and other chronic fatiguing conditions such as long COVID, to improve quality of life through symptom management.

Conversely, some studies reported no effects of pacing on ME/CFS symptoms [52], fatigue, physical functioning [35], or pain scores [49, 61]. Some studies even found pacing to have detrimental effects in those with ME/CFS, including a worsening of symptoms in 14% of survey participants recalling previous pacing experiences [52]. Furthermore, a worsening of fatigue [35, 59], and physical functioning from pre- to post-pacing [35, 57, 59, 61] was reported by the PACE trial and sub-analysis of the PACE trial [56, 57, 61]. The PACE trial [35], a large RCT (n = 639) comparing pacing with CBT and GET, reported GET and CBT were more effective for reducing ME/CFS-related fatigue and improving physical functioning than pacing. However, the methodology and conclusions from the PACE trial have been heavily criticised, mainly due to the authors lowering the thresholds they used to determine improvement [36–38, 67]. With this in mind, Sharpe et al. [56] surveyed 75% of the participants from the PACE trial 1-year post-intervention and reported pacing improved fatigue and physical functioning, with effects similar to CBT and GET.

All pacing intervention studies (5/5) implemented educational or coaching sessions. These educational components were poorly reported in terms of the specific content and how and where they had been developed, with unclear pedagogical approaches. Consequently, even where interventions reported reduction in PEM or improved symptoms, it would be impossible to transfer that research into practice, future studies, or clinical guidance, given the ambiguity of reporting. Sessions typically contained themes of pacing such as activity adjustment (decrease, break-up, and reschedule activities based on energy levels), activity consistency (maintaining a consistently low level of activity to prevent PEM), activity planning (planning activities and rest around available energy levels), and activity progression (slowly progressing activity once maintaining a steady baseline) [35, 48–51]. We feel it is pertinent to note here that although activity progression has been incorporated as a pacing strategy in these included studies, some view activity progression as a form of GET. The NICE definition of GET is “first establishing an individual's baseline of achievable exercise or physical activity, then making fixed incremental increases in the time spent being physically active” [34]. Thus, this form of pacing can also be considered a type of ‘long-term GET’ in which physical activity progression is performed over weeks or months with fixed incremental increases in time spent being physically.

Intervention studies attempted to create behaviour change, through educational programmes to modify physical activity, and plan behaviours. However, none of these studies detailed integrating any evidence-based theories of behaviour change [68] or reported using any frameworks to support behaviour change objectives. This is unfortunate since there is good evidence that theory-driven behaviour change interventions result in greater intervention effects [69]. Indeed, there is a large body of work regarding methods of behaviour change covering public health messaging, education, and intervention design, which has largely been ignored by the pacing literature. Interventions relied on subjective pacing (5/5 studies), with strategies including keeping an activity diary (3/5 studies) to identify links between activity and fatigue [35, 48, 50]. Given the high prevalence of ‘brain fog’ within ME/CFS [70–73], recall may be extremely difficult and there is significant potential for under-reporting. Other strategies included simply asking participants to estimate energy levels available for daily activities (2/5 studies [48, 51]). Again, this is subjective and relies on participants’ ability to recall previous consequences of the activity. Other methods of activity tracking and measuring energy availability, such as wearable technology [74–78] could provide a more objective measure of adherence and pacing strategy fidelity in future studies. Despite technology such as accelerometers being widely accessible since well-before the earliest interventional study included in this review (which was published in 2009), none of the interventional studies utilised objective activity tracking to track pacing and provide feedback to participants. One study considered accelerometery alongside an activity diary [51]. However, accelerometery was considered the outcome variable, to assess change in activity levels from pre- to post-intervention and was not part of the intervention itself (which was one pacing coaching sessions per week for 3 weeks). Moreover, most research-grade accelerometers cannot be used as part of the intervention since they have no ability to provide continuous feedback and must be retrieved by the research team in order to access any data. Consequently, their use is mostly limited to outcome assessments only. As pacing comprises a limit to physical activity to prevent push-crash cycles, it is an astonishing observation from this scoping review that only two studies objectively measured physical activity to quantify changes to activity as a result of pacing [51, 54]. If the aim of pacing is to reduce physical activity, or reduce variations in physical activity (i.e., push-crash cycles), only two studies have objectively quantified the effect pacing had on physical activity, so it is unclear whether pacing was successfully implemented in any of the other studies.

By exploring the pacing strategies previously used, in both intervention studies and more exploratory studies, we can identify and recommend approaches to improve symptoms of ME/CFS. These approaches can be categorised as follows: activity planning, activity consistency, activity progression, activity adjustment and staying within the Energy Envelope [50, 53, 60, 63]. Activity planning was identified as a particularly effective therapeutic strategy, resulting in improvement of mean scores of all symptoms included in the APQ-28, reducing current pain, improvement of physical fatigue, mental fatigue, self-efficacy, quality of life, and mental and physical functioning [50]. Activity planning aligns with the self-regulatory behaviour change technique ‘Action Planning’ [79] which is commonly used to increase physical activity behaviour. In the case of ME/CFS, activity planning is successfully used to minimise rather than increase physical activity bouts to prevent expending too much energy and avoid PEM. Activity consistency, meaning undertaking similar amounts of activity each day, was also associated with reduced levels of depression, exercise avoidance, and higher levels of physical function [63]. Activity progression was associated with higher levels of current pain. Activity adjustment associated with depression and avoidance, and lower levels of physical function [63]. Staying within the Energy Envelope was reported to reduce PEM severity [53, 60], improve physical functioning [53, 60] and ME/CFS symptom scores [53], and more hours engaged in activity than individuals with lower available energy [53]. These results suggest that effective pacing strategies would include activity planning, consistency, and energy management techniques while avoiding progression. This data is, of course, limited by the small number of mostly low-quality studies and should be interpreted with some caution. Nevertheless, these are considerations that repeatedly appear in the literature and, as such, warrant deeper investigation. In addition, and as outlined earlier, most studies are relatively old, and we urgently need better insight into how modern technologies, particularly longitudinal activity tracking and contemporaneous heart-rate feedback, might improve (or otherwise) adaptive pacing. Such longitudinal tracking would also enable activities and other behaviours (sleep, diet, stress) to be linked to bouts of PEM. Linking would enable a deeper insight into potential PEM triggers and mitigations that might be possible.

We feel it would be remiss of us to not specifically address the PACE trial within this manuscript, as five of the 17 included studies resulted from the PACE trial [35, 56, 57, 59, 61]. There has been considerable discussion around the PACE trial, which has been particularly divisive and controversial [37–39, 59, 67, 80, 81]. In the PACE trial, GET and CBT were deemed superior to pacing by the authors. Despite its size and funding, the PACE trial has received several published criticisms and rebuttals. Notably, NICE's most recent ME/CFS guideline update removed GET and CBT as suggested treatment options, which hitherto had been underpinned by the PACE findings. While we will not restate the criticisms and rebuttals here, what is not in doubt, is that the PACE trial has dominated discussions of pacing, representing almost a third of all the studies in this review. However, the trial results were published over a decade ago, with the study protocol devised almost two decades ago [82]. The intervening time has seen a revolution in the development of mobile and wearable technology and an ability to remotely track activity and provide real-time feedback in a way which was not available at that time. Furthermore, there has been no substantive research since the PACE trial that has attempted such work. Indeed, possibly driven by the reported lack of effect of pacing in the PACE trial, this review has demonstrated the dearth of progress and innovation in pacing research since its publication. Therefore, regardless of its findings or criticisms, the pacing implementation in the PACE trial is dated, and there is an urgent need for more technologically informed approaches to pacing research.

The first limitation to the literature included in this scoping review is that not all studies followed the minimum data set (MDS) of patient-reported outcome measures (PROMs) agreed upon by the British Association of CFS/ME Professionals (BACME) (fatigue, sleep quality, self-efficacy, pain/discomfort, anxiety/depression, mobility, activities of daily living, self-care, and illness severity) [83, 84]. All but one study included in this review measured illness severity, most studies included fatigue and pain/discomfort, and some studies included assessments of anxiety/depression. There was a lack of quantitative assessment of sleep quality, self-efficacy, mobility, activities of daily living, and self-care. Therefore, studies did not consistently capture the diverse nature of the symptoms experienced, with crucial domains missing from the analyses. The MDS of PROMs were established in 2012 [83, 84] and therefore, for studies published out prior to 2012, these are not applicable [35, 49, 51, 53, 54]. However, for the 12 studies carried out after this time, the MDS should have been considered elucidate the effects of pacing on ME/CFS. Importantly, despite PEM being a central characteristic of ME/CFS, only two studies included PEM as an outcome measure [55, 60]. This may be because of the difficulty of accurately measuring fluctuating symptoms, as PEM occurs multiple times over a period of months, and therefore pre- to post- studies and cross-sectional designs cannot adequately capture PEM incidence. Therefore, it is likely studies opted for measuring general fatigue instead. More appropriate longitudinal study designs are required to track PEM over time to capture a more representative picture of PEM patterns. Secondly, reporting of participant characteristics was inadequate, but in the studies that did describe participants, characteristics were congruent with the epidemiological literature and reporting of ME/CFS populations (i.e., 60–65% female) [85]. Therefore, in this respect, studies included herein were representative samples. However, the lack of reporting of participant characteristics limits inferences we can draw concerning any population-related effects (i.e. whether older, or male, or European, or people referred by a national health service would be more or less likely to respond positively to pacing). Thirdly, comparison groups (where included) were not ideal, with CBT or GET sometimes used as comparators to pacing [35], and often no true control group included. Penultimately, there is a distinct lack of high-quality RCTs (as mentioned throughout this manuscript). Finally, in reference to the previous section, inferences from the literature are dated and do not reflect the technological capabilities of 2023.

It is clear from the studies included in this scoping review for the last decade or more, progress and innovation in pacing research have been limited. This is unfortunate for several reasons. People with ME/CFS or long COVID are, of course, invested in their recovery. From our patient and public involvement (PPI) group engagement, it is clear many are ahead of the research and are using wearable technology to track steps, heart rate, and, in some cases, heart rate variability to improve their own pacing practice. While the lack of progress in the research means this is an understandable response by patients, it is also problematic. Without underpinning research, patients may make decisions based on an individual report of trial-and-error approaches given the lack of evidence-based guidance.

A more technologically-informed pacing approach could be implemented by integrating wearable trackers [77, 78, 86, 87] to provide participants with live updates on their activity and could be integrated with research-informed messaging aimed at supporting behaviour change, as has been trialled in other research areas [88–91]. However, more work is needed to evaluate how to incorporate wearable activity trackers and which metrics are most helpful.

A more technologically-informed approach could also be beneficial for longitudinal symptom tracking, particularly useful given the highly variable symptom loads of ME/CFS and episodic nature of PEM. This would overcome reliance on assessments at a single point in time (as the studies within this review conducted). Similarly, mobile health (mHealth) approaches also allow questionnaires to be digitised to make it easier for participants to complete if they find holding a pen or reading small font problematic [92]. Reminders and notifications can also be helpful for patients completing tasks [77, 93–95]. This approach has the added advantage of allowing contemporaneous data collection rather than relying on pre- to post-intervention designs limited by recall bias. Future work must try to leverage these approaches, as unless we collect large data sets on symptoms and behaviours (i.e. activity, diet, sleep, and pharmacology) in people with conditions like ME/CFS we will not be able to leverage emerging technologies such as AI and machine learning to improve the support and care for people with these debilitating conditions. The key areas for research outline in the NICE guidelines (2021 update) speaks to this, with specific mention of improved self-monitoring strategies, sleep strategies, and dietary strategies, all of which can be measured using mHealth approaches, in a scalable and labour-inexpensive way.

There is now an urgent public health need to address long COVID, with over 200 million sufferers worldwide [30]. Given the analogous symptomology between ME/CFS and long COVID, and the lack of promising treatment and management strategies in ME/CFS, pacing remains the only strategy for managing long COVID symptoms. This is concerning as the quality of evidence to support pacing is lacking. Given long COVID has reached pandemic proportions, scalable solutions will be required. In this context, we propose that technology should be harnessed to a) deliver, but also b) evaluate, pacing. We recently reported on a just-in-time adaptive intervention to increase physical activity during the pandemic [78]. However, this method could be adapted to decrease or maintain physical activity levels (i.e., pacing) in long COVID. This method has the advantage of scalability and remote data collection, reducing resource commitments and participant burden, essential for addressing a condition with so many sufferers.

This manuscript did not involve human participants, data, or tissues, so did not require ethical approval.

This paper does not contain any individual person’s data in any form.

We report no financial and non-financial competing interests.

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.