Definition and classification for adverse events following spinal and peripheral joint manipulation and mobilization: A scoping review

Introduction Spinal and peripheral joint manipulation and mobilization are interventions used by many healthcare providers to manage musculoskeletal conditions. Although there are many reports of adverse events (or undesirable outcomes) following such interventions, there is no common definition for an adverse event or clarity on any severity classification. This impedes advances of patient safety initiatives and practice. This scoping review mapped the evidence of adverse event definitions and classification systems following spinal and peripheral joint manipulation and mobilization for musculoskeletal conditions in adults. Methods An electronic search of the following databases was performed from inception to February 2021: MEDLINE, EMBASE, CINAHL, Scopus, AMED, ICL, PEDro, Cochrane Library, Open Grey and Open Theses and Dissertations. Studies including adults (18 to 65 years old) with a musculoskeletal condition receiving spinal or peripheral joint manipulation or mobilization and providing an adverse event definition and/or classification were included. All study designs of peer-reviewed publications were considered. Data from included studies were charted using a standardized data extraction form and synthesised using narrative analysis. Results From 8248 identified studies, 98 were included in the final synthesis. A direct definition for an adverse event and/or classification system was provided in 69 studies, while 29 provided an indirect definition and/or classification system. The most common descriptors to define an adverse event were causality, symptom severity, onset and duration. Twenty-three studies that provided a classification system described only the end anchors (e.g., mild/minor and/or serious) of the classification while 26 described multiple categories (e.g., moderate, severe). Conclusion A vast array of terms, definition and classification systems were identified. There is no one common definition or classification for adverse events following spinal and peripheral joint manipulation and mobilization. Findings support the urgent need for consensus on the terms, definition and classification system for adverse events related to these interventions.


Introduction
Spinal and peripheral joint manipulation and mobilization are interventions used by many healthcare providers to manage musculoskeletal conditions.Although there are many reports of adverse events (or undesirable outcomes) following such interventions, there is no common definition for an adverse event or clarity on any severity classification.This impedes advances of patient safety initiatives and practice.This scoping review mapped the evidence of adverse event definitions and classification systems following spinal and peripheral joint manipulation and mobilization for musculoskeletal conditions in adults.

Introduction
Spinal pain, including low back and neck pain, is the most common musculoskeletal problem globally, a leading cause of disability and absenteeism from work and is ever increasing [1].These factors contribute to increased socioeconomic burdens and costs [2].Clinical guidelines and best practice recommendations (e.g., NICE Guidelines) advocate the use of conservative interventions, including spinal and peripheral joint manipulation and mobilization, provided by a variety of healthcare professionals (e.g., chiropractors, naprapaths, osteopaths, physiotherapists, physicians, etc.) as a management option for uncomplicated presentations of spinal pain [3][4][5].Used globally by manual therapists as conservative interventions, spinal and peripheral joint manipulation involves the application of a high-velocity, low-amplitude force to a specific joint, whilst spinal and peripheral joint mobilization involves the application of a cyclic low-velocity force [6].
Similar to any medical intervention, joint manipulations and mobilizations are not without risk of harms or complications [7].Whilst serious harms have been reported to be rare [8][9][10][11], the consequences of such can be devastating, with considerable impact on those involved.Patient safety remains a top priority within healthcare, with a continued focus on preventing and minimising adverse events following any type of intervention [12,13].A 2015 North American Patient Safety Foundation expert panel emphasised the importance of patient safety as a public health issue with a main recommendation being the need for a common set of safety metrics for use across all practice settings, including primary or ambulatory care settings, which is where the majority of care is provided [14,15].
"Harms", "complications", "side-effects" and "adverse events" are among several commonly used terms in the literature describing undesirable outcomes of manual interventions (e.g., spinal and peripheral joint manipulation and mobilization), which are most commonly used to reduce pain and disability in patients with musculoskeletal complaints [16][17][18].Additionally, how these outcomes are defined and what constitutes an adverse event (or undesirable outcome) following spinal and peripheral joint manipulation and mobilization remains disparate [19][20][21].These outcomes may be further classified according to their severity (e.g., mild, moderate, severe), onset (e.g., during treatment, within 24-48 hours after treatment), duration (e.g., transient, short-lasting, permanent) or need for unplanned additional remedial or medical care (e.g., investigations, specialist referral, hospitalisation) [22,23].The kaleidoscope of domains and descriptors used in the literature to report and characterize these outcomes impedes attempts to advance patient safety initiatives and practices through a common and universal understanding of observed safety incidents.Although previous studies have highlighted this issue [16] and proposed frameworks for categorizing adverse events following manual therapy [23][24][25], there is still no standardization as to what constitutes an adverse event following such manual therapy interventions.A standardized and accepted adverse event typology would not only facilitate the development of strategies to minimise or prevent such events across all manual therapy professions that use these interventions, but more importantly, achieve consistency and precision in documenting and reporting such events.Specifically, an adverse events typology should include an operational definition of an adverse event so that identification, reporting and learning opportunities can be standardized across professions using spinal and peripheral joint manipulation and mobilization.
For these reasons, a scoping review of the literature is required.Combining the published knowledge from different professions, healthcare settings and musculoskeletal conditions will elucidate the current landscape and true extent of the problem.Findings from this scoping review will provide the evidence needed to conduct further research and move towards a consensus on the topic of adverse events.Ultimately, enhancing patient safety practices for spinal and peripheral joint manipulation and mobilization.

Aim and objectives
This scoping review aimed to map the scientific literature defining adverse events and their respective classification systems following spinal or peripheral joint manipulation and mobilization for musculoskeletal conditions in an adult population.Specific objectives included: 1.To describe how adverse events following spinal and peripheral joint manipulation and mobilization have been defined in the literature; 2. To describe how adverse events following spinal and peripheral joint manipulation and mobilization have been classified in the literature.

Design
This scoping review followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) for transparency in reporting [26].
The protocol was registered at the Open Science Framework Registry (10.17605/OSF.IO/ UBX2D) and designed by an international, interprofessional team of manual therapists (chiropractors, osteopaths and physiotherapists) with relevant clinical and methodological expertise.
A scoping review was chosen as this study focuses on examining and clarifying definitions and classification systems for adverse events following spinal and peripheral joint manipulation and mobilization that are used in the literature [27].

Stages
This review was conducted in 5 stages: (1) identifying the research question; (2) identifying relevant studies; (3) study selection; (4) charting the data; and (5) collating, summarizing and reporting the results [28,29].The optional consultation exercise (step 6) was not included within the scope of this specific manuscript as the results will be used to inform an e-Delphi study [30].
Stage 1: Identifying the research question.How does the scientific literature define adverse events and their respective classification systems for events that occur following spinal or peripheral joint manipulation and mobilization for musculoskeletal conditions in an adult population?Stage 2: Identifying relevant studies.Information sources.The following databases were searched from inception to 12 th February 2021: MEDLINE, EMBASE, CINAHL, Scopus, AMED, ICL, PEDro and Cochrane Library.Grey literature using Open Grey and Open Access Theses and Dissertations (OATD) were also searched.
Search strategy.The search strategy was designed by the authors with the assistance of an experienced health sciences librarian.The initial search strategy (S1 Table ) was developed for Ovid MEDLINE using medical subject headings (MeSH) and text words.This was subsequently adapted to the syntax and subject headings of the other databases that were searched.
Eligibility criteria.Studies were identified by using the eligibility criteria outlined in Table 1.
For the purpose of this study, mobilization was defined as a manual therapy technique comprising a continuum of skilled passive movements that were applied at varying speeds and amplitudes to joints [6].Manipulation was defined as a passive, high velocity, low amplitude thrust applied to a joint complex within its anatomical limit (the range of motion of the joint complex in which active and passive motion occurs and not beyond the joint's anatomic limit) [6].The term "adverse event" was adopted as an umbrella term to reflect any undesirable effect of spinal or peripheral joint manipulation and mobilization where terms such as "harms", "complications", "side-effects", etc. have also been used in the literature [16][17][18].
Stage 3: Study selection.This stage was conducted in 2 phases with each phase starting with a pre-screening team meeting to discuss inclusion and exclusion criteria.Both phases ▪ Case reports that do not mention the terms "adverse event", "complication", "side effect", "adverse reaction", etc. or do not imply the condition being an adverse event to manipulation or mobilization https://doi.org/10.1371/journal.pone.0270671.t001 were performed using Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia), an online tool developed for systematic reviews by the Cochrane Collaboration that follows the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.Five reviewers screened the same 20 publications to ensure interscreener calibration with weekly meetings to resolve any conflicts.Specifically, in phase 1, titles and abstracts were independently screened by two of the five reviewers to identify potentially relevant studies.Any disagreements were resolved through discussion.Phase 2 was based on the full texts of all studies identified as potentially relevant during phase 1.Similar to phase 1, two of the five reviewers independently screened full texts and any disagreements were resolved by discussion and consensus with all five reviewers.Stage 4: Charting the data.Data were extracted by all five reviewers working as a group using a standardized data extraction form that was first piloted with 20 included studies.All data extracted were checked for accuracy by two reviewers (MF and LG).Disagreements were resolved through discussion and consensus.
Extracted data included study characteristics (first author, year of publication, title, country, study design [original studies: case report/series, observational studies, consensus, survey, experimental trials, other designs; and clinical practice guidelines, review or study protocols]), participant population for original studies or study protocol (sample size, condition being treated [spinal, peripheral, mixed, unknown, not applicable], condition severity, condition chronicity [acute (<3 weeks), subacute (>3 weeks, <3 months), chronic (>3 months), mixed, unknown]), studies description for clinical practice guidelines and review (number of included studies, design of included studies), intervention characteristics (setting [community-based clinic/office, hospital, research clinic, academic institute, mixed], profession [chiropractic, naprapathy, osteopathy, physiotherapy, mixed, other, unknown], intervention [manipulation, mobilization, mixed]), and adverse event characteristics (definition, classification system [e.g., minor-moderate-major; mild-moderate-severe-serious], citations for adverse event definitions or classification systems, and whether the provided adverse event definition was direct (a clear statement of what was considered an adverse event) or indirect (indicated what was considered an adverse event without a clear statement [e.g., provided the question asked to participants during the study])).
Stage 5: Collating, summarizing and reporting the results.Extracted data were categorized into two groups: i) studies providing a direct definition and/or classification system for adverse events following spinal and peripheral joint manipulation or mobilization; and ii) studies providing an indirect definition and/or classification system for adverse events following these interventions.Specifically, studies providing a direct definition and/or classification were those that provided a clear statement of the study's operational definition and/or classification system for adverse events.For example: "For the purposes of this study, we adopted the following definition (derived from the ICH Guidelines for Good Clinical Practice): An adverse event (AE) can be any unfavourable and unintended sign, symptom, or disease temporally associated with the use of an intervention (treatment), which does not necessarily have a causal relationship with such treatment" [31].Studies providing an indirect definition and/or classification were those that did not provide a clear statement of their operational definition and/or classification system, but indicated what was considered an adverse event, for example, by providing the question used to collect adverse events in a survey.For example: "[. ..] possible adverse effects were assessed by 2 open-ended questions: (1) "Did your symptoms get worse after this treatment?"and (2) "Are you feeling any different symptoms after this treatment?"[32].A descriptive summary detailing the overall number of studies included in the review, their study characteristics as well as the data regarding adverse event definition and classification system extracted from included studies are provided.

Study selection
Electronic searches identified 8248 citations that resulted in 3963 unique citations to be screened for inclusion following the removal of duplicates.The titles and abstracts were assessed for their relevance to the review based on the eligibility criteria (phase 1 screening), where 3400 citations were excluded, resulting in 563 citations for full text review.The phase 2 screening excluded 465 full texts: 320 did not provide an adverse event definition or classification, 59 were not peer-reviewed publications, 20 included the wrong intervention (e.g., did not use joint manipulation or mobilization), 18 were conference proceedings, 15 included the wrong population (e.g., participants younger than 18 or older than 65 years old), 6 were not written in English, Portuguese or Italian and 27 were excluded for other reasons (e.g., full text not available, professional issue papers, etc.).As such, 98 studies were included in this scoping review (Fig 1 ).

Data synthesis
Studies with a direct adverse event definition and/or classification system.The studies that provided a direct definition of an adverse event and/or classification system (n = 69) were published between 1993 and 2020 and were mostly from North America (n = 28) and Europe (n = 27).Twenty-two studies focused on joint manipulation, 4 focused on mobilization and 43 included both techniques.Most were conducted in an academic institute (n = 26) and provided the intervention to the spine (n = 48).
In addition to the term "adverse events", the term "side-effects" was used in 4 studies, "sequelae" was used in 4 studies, "complication" in 3 studies, "incident" in 2 studies and one study used the term "reaction".New or worsened complaints or symptoms were described as adverse events in 18 studies.Twenty studies described adverse events as "unpleasant", "unfavourable", "unintended", "unexpected" and/or "undesired responses", and 5 studies as "untoward medical occurrences".Common descriptors composing the adverse event definition statements were identified in the studies that provided a direct definition of an adverse event (n = 49).Specifically, causality was incorporated in the definition provided by 21 studies (e.g., "Adverse events are unexpected events [. ..] without evidence of causality" [33]).Symptom severity was used to define adverse events in 20 studies (e.g., "Adverse events are moderate to severe [. ..]" [34]).Symptom onset was included in the definition provided by 19 studies (e.g., "Adverse events were events that occurred within 24 hours following the treatment" [19]).Symptom duration was used as a descriptor in 10 studies (e.g., "[. ..] increased pain and/or

Study Design Definition Classification system
Powell et al. 1993 [8] Case Report / series The definition of "complication" is broadened to include indirect injury in which treatment postpones the appropriate diagnosis or continues in spite of progressive symptoms or signs, then the incidence may be higher.
Severe: if the treatments were unsuccessful, resulted in major disability or death.
Senstad et al. 1996a [36] Survey Some type of unpleasant side effect after the previous treatment Adequate reactions: does not influence the working ability with spontaneous remission completed, at the latest, 2 days after the manipulation Exceeding reactions: objective worsening of the pre-existing state, with decreased work capacity, and spontaneous remission exceeding 2 days

Authors used:
On a scale from 1 to 4: Leboeuf-Yde et al. 1997 [37] Survey Any additional discomfort after the previous treatment Common/Normal reactions: early onset, mild-to-moderate severity and gone within 48 hours

Self-reported unpleasant reactions
Light or moderate reactions: little or no effect on normal daily activities

Thiel et al. 2006 [38] Survey
A patient safety incident can be defined by: Low Harm: Incident required extra observation or minor treatment (additional therapy or medication over short period of time, does not include admission to hospital or attending as an outpatient on repeated occasions) Moderate Harm: Incident resulted in a moderate increase in additional treatment (admission to hospital or attending as an outpatient on repeated occasions, or requiring surgery or prolonged episodes of care) and caused significant but not permanent harm ''That was a threat to my patient's well-being, and I don't want it to happen again" Severe Harm: Incident resulted in permanent harm (harm or disability directly related to incident and leading to permanent lessening of bodily functions, or sensory, motor, physiologic or intellectual deficit)

Death: Incident directly resulted in death
Near Miss: No harm occurred-Incident had potential to cause harm but was prevented either by chance or deliberate action

Rubinstein et al. 2007 [9]
Observational Study A new related complaint that was not present at baseline or a worsening of the presenting complaint or an existing complaint by >30% compared with baseline Intense AE: any AE fulfilling our definition of an adverse event and that also scored �8 in intensity on the 11-point NRS.
Serious AE: events resulting in death, life-threatening situations, the need for admittance to a hospital, or temporary or permanent disability.

Rubinstein et al. 2008 [39]
Observational Study A new related complaint that was not present at baseline, or a worsening of the neck pain or any other existing complaint by >30% compared to baseline.
Intense AE: any AE fulfilling our definition of an adverse event, and which also scored �8 in intensity on the 11-point NRS.
Serious AE: events resulting in death, life-threatening situations, the need for admittance to a hospital, or temporary or permanent disability.

Rubinstein et al.
2008 [40] Observational Study Either a new complaint or the worsening of an existing complaint by more than 30% compared to baseline (more than MCID on 11-point NRS) Intense AE: any AE fulfilling the definition, and which also scored 8 or higher in intensity on the 11-point NRS.
Serious AE: events resulting in death, life-threatening situations, the need for admittance to a hospital, or temporary or permanent disability.
(Continued )  [35] Case Report / series Side-effects of a benign nature included increased pain and/ or stiffness of short duration Mild to moderate: little or no influence on daily activities, brief, with spontaneous recovery and lasting no more than a few days.

Major complications: irreversible
Wangler et al. 2011 [41] Consensus An adverse event is an injury related to medical management, in contrast to a complication of disease.
Preventable or Non-preventable.
Adverse drug reaction is a complication that occurs when the medication is used as directed and in the usual dosage.
Preventable AE: An adverse event caused by an error or other type of systems or equipment failure.
An adverse event is the result of a care delivery problem related to chiropractic management, in contrast to complications of disease.

Survey
Harm is any negative health outcome observed in a clinical trial, regardless of whether it was definitively related to the intervention.This is true even when the harm does not become apparent until the treatment has been used by many patients for extended periods.This term, therefore, incorporates concepts related to adverse events, adverse reactions, side effects, and safety, and favors disclosure regardless of causation.Observational Study

Benign
Any unfavorable and unintended sign, symptom, or disease temporally associated with the use of a medical treatment or procedure that may or may not be considered related to the medical treatment or procedure.
Serious AE: result in death or hospitalization, or caused permanent disability.

Authors used:
Responses of worse or much worse were considered to indicate that an adverse event had occurred Exacerbation of the patients' chief complaints or a new symptom.

Do et al. 2018 [53]
Protocol Undesirable and unintentional signs (i.e., abnormal laboratory test results), symptoms, or disease occurring after treatment, which does not have to have a causal relationship with the treatment Mild: no need for additional procedures and no great interference with the subject's everyday life (function) Moderate: significant interference of the subject's everyday life (function), probable need for additional procedures but followed by resolution after the procedure Severe: calling for advanced procedure, and leaving sequelae will be applied for evaluation.
Tabell et al. 2019 [54] Observational Study Pain and loss of function with impact on daily living or work On a 0-10 NRS scale: Survey Any unfavourable and unintended sign, symptom, or disease temporally associated with the use of an intervention (treatment), which does not necessarily have a causal relationship with such treatment.

Puentedura et al.
2015 [66] Review The sequelae following thrust joint manipulation to the spine that are medium to long term in duration, with moderate to severe symptoms, and of a nature that is serious, distressing and unacceptable to the patient and requires further treatment Unwanted side effect: short term, mild in nature, non-serious, transient and reversible consequences of the treatment (Continued ) stiffness of short duration" [35]).In general, studies including chiropractors defined adverse events as an "unpleasant reaction" or "new or worsened symptom" more often than studies including other professions.Among the studies that provided a classification system (n = 50), 23 only described the end anchors of the classification (e.g., mild/minor and/or serious), 26 provided description of additional classification categories (e.g., moderate, severe, etc.) and 3 described a classification system not including severity (e.g., common and uncommon; preventable and not preventable).Common domains that were used to describe the severity classification categories included: intensity (e.g., "We classified adverse event intensity as NRS score of 1 to 3 = mild, score of 4 to 6 = moderate; and NRS score of 7 to 10 = severe" [47]); duration (e.g., mild = less than 24 hours, moderate = between 24 hours and 1 week, major = over 1 week [24]); functional impact (e.g., mild = function intact, moderate = function modified, major = function impaired [79]); and requirement of additional treatment (e.g., mild = no additional intervention, moderate = may need additional intervention, severe = required intensive intervention [55]).Overall, studies conducted in Asia classified adverse events according to the Spilker classification [90] more often than those conducted in other regions.
Among the 69 studies that provided a direct definition and/or classification system, 56 cited a reference or a source for the definition used, while 13 did not provide any reference or source.In total, 78 unique references were cited of which 55 were peer-reviewed publications

Author, year, citation Study Design Classification system
Stern et al. 1995 [73] Case Report / series A minor complication was defined as increasing symptoms as reported by the patient.
A major complication was defined as emergency surgery after the treatment.

Moderate: reversible and serious
Major / serious: irreversible

UK BEAM Trial
Team 2004 [75] Trial Serious adverse events: treatment related events leading to hospital admission or death within one week

Gibbons et al. 2006 [76]
Other design Transient: begin within 4 hours of receiving treatment and typically resolve within the next 24 hours.

Substantive reversible impairment
Serious non-reversible impairment Thiel et al. 2007 [77] Survey Minor adverse events: worsening of presenting symptoms or onset of new symptoms, immediately, and up to 7 days, after treatment Significant (serious) event: referred to hospital and/or severe onset/ worsening of symptoms immediately after treatment and/or resulted in persistent or significant disability/incapacity.

Haneline et al. 2009 [78]
Observational Study Serious adverse events: events resulting in death, life-threatening situations, the need for admittance to a hospital, or temporary or permanent disability.

Carnes et al. 2010a [23]
Consensus 'Mild' and 'not adverse' adverse events: short term and mild, they are non-serious, the patient's function remains intact, they are transient/reversible and no treatment alterations are required because the consequences are short term and contained.
'Moderate' adverse events: the same as 'major' adverse events but only moderate in severity.
'Major' adverse events: medium to long term, moderate to severe and unacceptable; they normally require further treatment and are serious and distressing.

Carlesso et al. 2011 [79]
Other design Not Adverse: short term duration, acceptable severity, intact function and no other explanation possible Mild: short term (hours to 2 days) duration, 0. Other design Serious adverse events: symptoms with immediate onset after treatment and with persistent or significant disability.(Continued ) and 23 were books, websites, online documents or other sources (Table 5).While most of the references were related to manual therapy (n = 46), studies and sources from other areas (e.g., oncology, pharmacology) and organizations (e.g., Agency for Healthcare Research and Quality, World Health Organization, National Patient Safety Agency) were also referenced (n = 29).Generally, studies including chiropractors cited Senstad et al. (1996aSenstad et al. ( , 1996bSenstad et al. ( , 1997) [36,57,91] ).
Studies with an indirect adverse event definition and/or classification system.The studies that provided an indirect definition of an adverse event and/or classification system (n = 29) were published between 1996 and 2021, mostly from North America (n = 14) and The United Kingdom (n = 6).Eleven studies focused on spinal and/or peripheral joint manipulation, 1 focused on mobilization and 17 included both techniques.Most were conducted in an academic institute (n = 14) and provided manipulation or mobilization to the spine (n = 19).Frydman et al. 2018 [84] Protocol Serious adverse events are defined as events that are fatal, lifethreatening or lead to hospitalisation.

Smith et al. 2019 [85]
Review Serious adverse events associated with manipulative therapies are defined as conditions that lead to hospital admission, permanent damage or death.
Yao et al. 2020 [86] Trial Nonserious adverse events: self-limited, and no permanent injuries occurred Serious adverse events: caused death, were life threatening, or necessitated admission to the hospital
Among the 29 studies that provided an indirect definition and/or classification system, 14 cited a reference or a source for the definition used, while 15 did not.In total, 27 unique references were cited among which 24 were peer-reviewed publications and 3 were books, websites, online documents or other sources (Table 9).Most of the references were related to manual therapy, including spinal and peripheral joint manipulation and mobilization, (n = 20); however, other areas (e.g., aviation) and sources (e.g., clinical practice guideline, dictionary) were also cited (n = 7).

Discussion
This study mapped the scientific literature discussing the definition of adverse events and their classification systems following spinal and peripheral joint manipulation and mobilization for musculoskeletal conditions in an adult population.Based on the 98 included studies, heterogeneous adverse event definitions and classification systems were identified.This empirical evidence of heterogeneity highlights the need for international and interprofessional consensus on a standardized definition and classification system so that patient safety practices for spinal and peripheral joint manipulation and mobilization can be more homogeneous, facilitating synthesis of findings and outcomes and, consequently, improving patient care.

Definition components
Despite adverse events following spinal and peripheral joint manipulation and mobilization being investigated since the 1990s, a clear definition and classification system has yet to be

Study design Definition
Giles et al. 2003 [98] Observational Study Events requiring remedial treatment; minor transient increase in spinal pain symptoms that did not qualify as an adverse event Hay et al. 2005 [94] Trial "An exacerbation of pain after the initial assessment'' de Oliveira et al. 2013 [32] Trial "Did your symptoms get worse after this treatment?" "Are you feeling any different symptom after this treatment?"Satpute et al. 2020 [99] Case Reports/ Series "There were no adverse events during or after each treatment session in terms of dizziness, increase in pain intensity, or induction of other symptoms.''https://doi.org/10.1371/journal.pone.0270671.t006 established.Our findings highlight that even the terms used to refer to adverse events vary, ranging from "side effects" to "symptomatic responses" to "harms", "complications", "adverse response/reaction/effects/events/ experience", among others [36,42,45,48,66,71,91].Given the importance of this topic and the increased focus of healthcare on patient safety [13], it is surprising that the standardization of terms, definition and classification system have not yet been established.This, in turn, could have contributed to the lack of adverse event reporting systems within professions that use joint manipulation and mobilization interventions.An established reporting system that accurately collects the number of spinal and peripheral joint manipulations and mobilizations delivered [120] would allow more precise estimations of the incidence of adverse events following these interventions and potential contributing factors to such events could be investigated.Our findings identified descriptors commonly used to define adverse events and classification systems: causality, symptom severity, onset and duration, and action taken.Establishing

Clinical Practice Guideline
Non-treatment adverse events-not associated with a treatment modality, but that occur in the clinical setting; Unforseen-treatment adverse events-associated with a treatment modality, but not a known or observable risk factor; Forseen-treatment adverse events-associated with a treatment modality and predicted by an observable risk factor

Survey
Variation in terms of frequency and severity, ranging from the more frequent minor/benign adverse events to rare and serious adverse events Zhang et al. 2021 [108] Protocol "Shedding criteria: 1) intolerable adverse reactions; 2) serious adverse reactions; 3) the patients' pain continued to increase, which proved that trial participation was not suitable; 4) the patient's health may be damaged (for example, serious complications)'' NRS: numeric rating scale; SMT = spinal manipulative therapy. https://doi.org/10.1371/journal.pone.0270671.t007 causality between the delivery of a manual intervention and an adverse event is challenging.This relationship/association has long been a discussion within manual therapy [23,121,122], other healthcare arenas [123][124][125][126][127] and the overall field of patient safety [18,128].Recently, a

Survey
Benign and short-lasting discomfort "The patient was asked for the degree of pain (four choices ranging from "minor" to "unbearable").The chiropractor then asked three questions: 1) "How soon after the treatment did the reaction start?", 2) "How long did the reaction last?", and 3) "Could you carry out your usual work?" '' tool to assess causality of adverse events associated with any therapeutic interventions has been developed that could assist with this complex issue [129].In addition to being one of many components of adverse event definitions identified in this study, this standardized tool to assess causality may assist with the identification of predisposing factors which, in turn, could contribute to the development of prevention and mitigation strategies of similar adverse events [130].Therefore, a standardized operational definition for what constitutes an adverse event is needed to allow for the identification of adverse events across professions and this tool could be used to assess the causality of accurately identified adverse events.

Cited by # studies
Symptom severity (or intensity), onset and duration, and action taken (e.g., medication use, seeking unplanned medical care) were descriptors commonly used to define both what constitutes an adverse event and describe the classification system categories (Tables 2-4).For example, symptom severity was observed within adverse event definitions when it included the worsening and aggravation of a symptom (e.g., increased pain).Additionally, adverse events were classified as minor (or mild), moderate or major (or severe) based on the symptom duration, with minor (or mild) adverse events being short-duration (e.g., less than 24 hours) and major (or severe) adverse events being long term.Although different studies used different thresholds (e.g., considering short duration 24 or 48 hours), these descriptors were observed in most included studies.
Overall, in healthcare, adverse events have been classified based on the intervention (e.g., vaccine adverse event), anatomical location (e.g., eye adverse event), severity (e.g., serious adverse event), or causality (e.g., causal adverse event) [18].However, specific classifications within medical areas have been developed to better characterize adverse events, contributing to advancements within patient safety by enabling better synthesis of information [131][132][133][134].For example, Kaafarani et al. (2014) proposed that the classification of intraoperative adverse events to range from Class I (injury requiring no repair with the same procedure) to Class VI (intraoperative) [134].Kaufman (2016) identified that not all adverse drug reactions fit into the previously established types A (predictable) and B (novel responses) and proposed the addition of type C (continuing), type D (delayed use) and type E (end of use) reactions [133].Therefore, developing a standardized classification specific for adverse events following spinal and peripheral joint manipulation and mobilization could provide a common language for all professions that use these interventions and facilitate identification, reporting and communication about adverse events, promoting interprofessional learning and contributing to advancing patient safety.

Multidisciplinarity
Although adverse events following spinal and peripheral joint manipulation and mobilization have been the focus of several studies, these often include one profession (e.g., chiropractic, naprapathy, osteopathy, physiotherapy, etc.) [19,31,47,60].Given the number of professions using these manual therapy interventions, it is possible that the inter-professional knowledge exchange related to definitions and classification of adverse events was limited as each profession focused on their individual (siloed) professional communities rather then the intervention at large.Indeed, this review identified that included studies tended to cite references that were published by authors in the same profession.Although communication across health-related professions has been observed to be well-established and a common practice among academic communities [135,136], joint manipulation and mobilization providers have been described to present an unique culture related to patient safety [137].Specifically, divergent intra-and inter-profession beliefs, overlapping scopes of practice and perceived business competition may prevent interprofessional communications focused on adverse events following these interventions [137].However, in order to advance joint manipulation and mobilization safety initiatives, enhanced interprofessional communications and collaborations are not only possible but fundamental.We have attempted to address this issue by establishing an international, multidisciplinary working group investigating adverse event definition and classification systems across all professions using joint manipulation and mobilization.Another example includes the international framework for risk assessment of cervical artery dysfunction [138], which included a multidisciplinary research team.Furthermore, in 2010a, Carnes and colleagues conducted a multidisciplinary Delphi study with the aim to seek an expert consensus definition for adverse events applicable to all professions that use manual therapy [23].Similarly, Carlesso et al. (2011) explored how patients receiving manual therapy from different professions defined adverse events [79].These are two of the few multidisciplinary studies, including different professions (i.e., chiropractic, osteopathy, physiotherapy) that use spinal and peripheral joint manipulation and mobilization and were referenced by 19% and 7% of the 98 included studies in this scoping review, respectively.This percentage of referencing is slightly lower than the average 20%-35% interprofessional referencing in medical sciences [136] and could potentially explain, at least partially, the heterogenous adverse event definition and classification systems observed in this study.Although Carnes et al. (2010a) did not achieve consensus on a succinct adverse event definition, a proposed classification system was clearly determined and described [23].Remarkably, even though this work was published over a decade ago, the definition of an adverse event and their classification systems remained noticeably heterogeneous in the manual therapy literature, including in studies published after Carnes et al. (2010a).This reinforces the possibility of limited interprofessional knowledge exchange related to this specific topic, and the importance of the broad dissemination of results going beyond individual professions, as well as efforts from all professions to enhance interprofessional, topic-related knowledge, rather than profession-specific.

Geography
The heterogeneity in adverse event definitions and classification systems identified in this study could potentially be due to the fact that spinal and peripheral joint manipulation and mobilization are interventions commonly used by different professionals located in different geographical locations [139][140][141].Variation in professions' scope of practice and regulations between continents, countries and even regional jurisdictions could lead providers to use their own definition for an adverse event, based on their local practices and regulations [139,142].Additionally, the emergence of litigation most commonly related to serious adverse events following manipulation leading to significant disability, such as vertebral artery dissection, cauda equina syndrome, etc., may have contributed to the development of local definitions [11,21,80,104].As the number of serious and life-threatening adverse events reports following manipulation increased, so did the number of malpractice lawsuits against professionals who use these interventions [143][144][145].Consequently, legal courts, lawyers and malpractice insurers were likely compelled to develop local definitions in order to process and rule on such cases.Given that any practising provider is vulnerable to experiencing malpractice lawsuits against them, they may feel bound to these local definitions to be consistent with the environment in which they practice.
Additionally, given the diverse geographical locations in which spinal and peripheral joint manipulation and mobilization are used, cultural differences and their influence on individual beliefs and behaviours could also be a potential contributor to the adverse event definition and classification system heterogeneity found in this study [146].Culture refers to values, norms, and codes that collectively shape the beliefs, attitudes, and behavior of a group [147].Indeed, the impact of culture on health has been widely investigated as better understanding cultural contexts advances the knowledge of inter-personal roles, connections, and relationships (whether positive or negative), as well as allowing the understanding of how individuals are shaped and their health [147][148][149][150]. Consequently, cultural differences can play an important role in how adverse events after these interventions are defined and classified and may have a significant contribution to the heterogeneity identified in this study.This review identified trends in citations where specific continents used specific references more often in comparison to other continents.Besides demonstrating a potential limitation in knowledge exchange across geographical locations, this finding highlights the paucity of studies related to this topic from some parts of the world, including Africa and South America.Therefore, including these continents when developing a standardized adverse event definition and classification system is of great importance not only to take into consideration geographical and cultural particularities, but also to support the development of investigations related to this topic in these locations.

Future studies
Based on these findings, an e-Delphi study will be conducted to establish a standardized adverse event definition and classification system that can be prospectively used across multiple professions [30].This has the potential to greatly advance patient safety as it would provide a standardized framework for data to be collected and synthesized in an uniform manner.This would then provide all stakeholders of spinal and peripheral joint manipulation and mobilization interventions a comprehensive patient safety profile for the adult population with musculoskeletal conditions.Insights gained from this profile could assist with the formation and streamlining of clinical guidelines and further research capacities.

Strengths and limitations
Strengths of this study include the involvement of an interprofessional research group with clinical and methodological expertise, and development of the protocol a priori for transparency.Additionally, this review was not limited by country or profession; therefore, our findings are representative and transferable to an international and interprofessional audience.
Although the search included several potential terms related to adverse events, it is possible that potentially relevant studies that used alternate terms to describe adverse events were not captured.The search was also limited to studies published in English, Italian and Portuguese languages; potentially relevant studies published in other languages (such as German, French, Dutch, etc.) were not captured.Additionally, adverse event definitions provided by included studies were categorized into "direct" and "indirect".Although this categorization was clearly defined (i.e., direct definition provided a clear statement of what was considered an adverse event; indirect definition indicated what was considered an adverse event without a clear statement [e.g., provided the question asked to participants during the study]), it is not an established categorization and contains some level of subjectivity.

Conclusion
Findings identified that a vast array of terms, definitions and classification systems for adverse events following spinal and peripheral joint manipulation and mobilization have been published.Within this array of literature, there was no one standardized adverse event definition or classification system for adverse events following these interventions that is commonplace and widely used.This suggests that establishing a consensus on standardized terms, definitions and classification systems for adverse events related to these interventions is urgently needed and could advance strategies to enhance patient safety for all professions who deliver these interventions.

Table 2 .
(Continued) : mild to moderate intensity, and resolve spontaneously Minor: short term, mild intensity and do not require treatment; gets better on their own in a few days Moderate: medium to long term and of moderate intensity Major: medium to long term, require treatment and are of moderate to severe intensity Serious: result in death, hospitalization, or permanent injury Definitions based on duration and/or severity of the reaction: Short minor: NRS �3 and <24 hours of duration Long minor: NRS �3 and �24 hours of duration Short moderate: NRS >3 and <24 hours of duration Long moderate: NRS >3 and �24 hours of duration Mild AE: 1-3/10 on an 11-point NRS Moderate AE: Original pain increased regionally or/and radically after the intervention, with its intensity rated below 8/10 on an 11-point NRS (From questionnaire [additional file 1] = 4-6/10 on 11-point NRS) Any additional problems or increasing difficulties with the activities of daily living after spinal manipulation Severe AE: the radicular pain intensity is rated above 8/10 on an 11-point NRS, ankle-foot sensorimotor function

Table 2 .
(Continued)Mild or benign events: short-term, non-serious, the patient's function remains intact, and they are transient/reversible; no treatment alterations are required because the consequences are short-term and contained.

Table 5 . References cited by studies providing a direct adverse event definition and/or classification system (n = 78). Cited by # studies Reference Peer-reviewed articles (n = 55) Books, websites and other sources (n = 23)Table 5 .
(Continued) ProtocolSerious adverse events: any unfavorable sign, symptom, or disease temporally associated with the treatment, whether or not caused by the treatment that results in death or is life-threatening or results in inpatient hospitalization or prolongation of existing hospitalization for more than 24 hours with a persistent or significant incapacity or substantial disruption of the ability to conduct normal life functions.NRS: numeric rating scale.https://doi.org/10.1371/journal.pone.0270671.t004

Table 7 . Studies providing indirect adverse event classification system only (n = 10). Authors, year, citation Study design Classification system Barrett et al. 2000
[100]Survey "Hardly any discomfort, mild discomfort, moderate discomfort, severe discomfort, worst possible discomfort-one hour, one day and two days after SMT''