Characterizing Adverse Events Reported Immediately After Osteopathic Manipulative Treatment

Methods

In the current study, the dataset was collected during a prospective, longitudinal, multicenter observational study evaluating the use of OMT and its resultant outcomes based on patient and clinician report. The study protocol was approved by appropriate local institutional review boards and registered with ClinicalTrials.gov (#NCT02395965). All patients provided informed consent prior to participating.

Patients aged 18 years or older receiving OMT from a DO-Touch.NET member clinician (osteopathic physicians, allopathic physicians, and foreign-trained osteopaths who use osteopathic manipulative procedures) were eligible to participate in the study. Patients who had difficulties communicating in English, dementia, or psychological conditions that could affect their ability to provide accurate information were excluded. Patients and clinicians provided data through a series of questionnaires that were administered in electronic format or paper format, depending on patient preference. Electronic data were entered by patients directly into the data capture tools: Assessment Center (Chicago, IL) was used from 2011-2012, and REDCap⁸ (Nashville, TN), hosted at A.T. Still University, was used from 2013-2014. For the paper format of the survey, patients were provided postage-paid envelopes and were asked to mail the completed forms to the DO-Touch.NET coordinating center; data from the paper forms were entered twice into the electronic format database by DO-Touch.NET staff to ensure data quality. Background and demographic data on patients and clinicians, data collected during and immediately after an office visit where OMT was provided, and data collected on a daily basis for 7 days after OMT populated this dataset. For the current study, data were analyzed from patient responses to questionnaires administered immediately before and after OMT and the medical record for the office visit.

The questionnaire administered immediately after OMT, consisting of a 5-point ordinal rating scale (much better, better, about the same, worse, much worse) assessing how the patients felt compared with before their OMT session, was completed after the clinician left the treatment room. For those patients who indicated they felt their condition had changed, either for better or worse, a follow-up, open-ended question asked them to describe how it had changed. Responses of worse or much worse were considered to indicate that an adverse event had occurred. Two reviewers (B.F.D., W.J.B.) independently coded the types of adverse events experienced and whether the adverse event was an exacerbation of the patients’ chief complaints or a new symptom based on the descriptions provided by the patients. A maximum of 3 chief complaints were defined by patients before receiving OMT. Discrepancies between the 2 reviewers were resolved through discussions between them until consensus was reached.

Information regarding the OMT techniques used was extracted from medical records. Documented treatment techniques were high-velocity, low-amplitude (HVLA); counterstrain/facilitated positional release; muscle energy; articulatory/Still; myofascial release; soft tissue; visceral; cranial; indirect/functional; and balanced ligamentous tension/ligamentous articular strain techniques.

The incidence rates and 95% CIs for the types of adverse events reported immediately after OMT were calculated using generalized logistic regression models, in which the patient was treated as a random effect to account for the dependence between office visits for the same patient. Generalized logistic regression models were also used to test for associations of adverse events with demographic characteristics (sex, race/ethnicity, and age) and with individual OMT techniques after accounting for demographic characteristics, and to calculate the corresponding adjusted ORs with 95% CIs. P values less than .05 were considered statistically significant. The statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc).

Results

A total of 925 patients participated in the study at 1915 office visits to 43 clinicians (41 osteopathic physicians, 1 allopathic physician, and 1 Canadian-trained osteopath). Of the 884 patients (96%) who provided data immediately after OMT at 1847 office visits, 663 (76%) were women (7 chose not to provide their gender); 794 (92%) identified as white, 38 (4%) as Hispanic or Latino/a, 10 (1%) as multiple races, 6 (1%) each as Asian and as American Indian or Alaskan Native, 5 (1%) as black or African American, and 1 (<1%) as Native Hawaiian or Pacific Islander (24 chose not to provide their race/ethnicity); and the mean (SD) age was 51.8 (15.8) years with a range of 18 to 98 years (22 chose not to provide their age). The number of office visits per patient ranged from 1 to 36, with 615 patients (70%) contributing data from only 1 office visit. Patients at 39 office visits (2.1%) reported feeling worse immediately after OMT and at 6 (0.3%) reported feeling much worse, indicating an adverse event had occurred; patients at 1595 (86.4%) reported feeling better or much better (Figure 1). The 45 office visits where adverse events occurred were reported by 27 patients; 21 patients reported adverse events at 1 office visit, 4 at 2 office visits, 1 at 3 office visits, and 1 at 13 office visits.

Figure 1.

Patient-reported responses immediately after osteopathic manipulative treatment (N=1847 office visits).

For 20 (44%) of the office visits with adverse events, patients did not provide information characterizing how they felt worse. Thus, the reviewers were unable to determine the type of adverse event and whether it was an exacerbation of the patients’ chief complaints or a new symptom. Adverse events that were exacerbations of the chief complaints were identified for 8 office visits (18%), and adverse events that were different from the reported chief complaints were identified for 10 office visits (22%). For 14 office visits (31%), the coders identified the type of adverse event but could not determine whether the adverse event was an exacerbation of the chief complaints.

The incidence of adverse events occurring immediately after OMT was estimated at 2.5% (95% CI, 1.3%-4.7%), a rate of 25 adverse events per 1000 treatments (Figure 2). Pain/discomfort was the most commonly identified type of adverse event (16 of 1847 office visits; incidence, 0.9%; 95% CI, 0.5%-1.6%). Women reported adverse events more frequently than men (OR, 13.9; 95% CI, 1.7-115.6; P=.01) (Table). There was no significant effect of race/ethnicity (OR, 0.6; 95% CI, 0.1-4.8; P=.61) or age (OR, 0.8; 95% CI, 0.5-1.3; P=.35).

Figure 2.

Incidence of types of adverse events identified as occurring immediately after osteopathic manipulative treatment (N=1847 office visits).

Patients who were treated with HVLA (adjusted OR, 0.22; P<.001), counterstrain/facilitated positional research (adjusted OR, 0.26; P=.03), muscle energy (adjusted OR, 0.29; P<.001), articulatory/Still (adjusted OR, 0.43; P=.03), and myofascial release (adjusted OR, 0.46; P=.02) were less likely to experience an adverse event immediately after OMT than those who were not treated with those techniques (Figure 3). Patients who were treated with balanced ligamentous tension/ligamentous articular strain (adjusted OR, 3.25; P=.02), indirect/functional (adjusted OR, 2.78; P=.006), and cranial (adjusted OR, 1.92; P=.03) had a higher likelihood of experiencing an adverse event. With the possible exception of HVLA, the distribution of the types of adverse events was similar for the 10 OMT techniques examined (Figure 4). While HVLA appeared to be associated with lower incidence of pain/discomfort than the other techniques, the sample sizes were too small to warrant statistical analysis.

Figure 3.

Association of osteopathic manipulative treatment techniques and incidence of adverse events reported immediately after osteopathic manipulative treatment. Abbreviations: BLT/LAS, balanced ligamentous tension/ligamentous articular strain; CS/FPR, counterstrain/facilitated positional release; HVLA, high-velocity, low-amplitude; MFR, myofascial release.

Figure 4.

Distribution of types of adverse events for individual osteopathic manipulative treatment techniques. Abbreviations: BLT/LAS, balanced ligamentous tension/ligamentous articular strain; CS/FPR, counterstrain/facilitated positional release; HVLA, high-velocity, low-amplitude; MFR, myofascial release.

Discussion

In the current study, patients at 2.4% of office visits reported feeling worse or much worse immediately after OMT. None of the reports were considered a serious adverse event, which would have resulted in death or hospitalization, or caused permanent disability. The types of adverse events found were milder with symptoms consistent with those previously reported in the manual therapy literature,^9-11 namely pain/discomfort, headache, tiredness/fatigue, numbness/tingling, nausea/vomiting, lightheadedness, and stiffness. The most commonly reported adverse event was an increase in pain/discomfort.

A majority of the current evidence on the characterization of adverse events after manual therapies is focused on chiropractic high-velocity thrust techniques.¹⁰ Case series reports indicated that 36 major adverse events from cervical spine manipulation occurred between 1995 and 2001, most commonly arterial dissection leading to stroke.¹² Other events included disk herniation, retinal artery occlusion, epidural hematoma, phrenic nerve paralysis, and death in 2 cases.¹² A systematic review of adverse events from spinal manipulation in the pediatric population reported 9 serious adverse events, consisting of significant neurologic and orthopedic complications and 2 deaths, as well as 2 moderate and 3 mild adverse events.¹³ Although larger studies did not identify serious adverse events following manual therapy, the incidence of mild to moderate adverse events is noteworthy. After manual therapy, 60.9% of 465 patients reported at least 1 posttreatment reaction, most commonly headache (19.8%), stiffness (19.5%), local discomfort (15.2%), radiating discomfort (12.1%), and fatigue (12.1%).¹⁴ Most of these reactions began within 4 hours of treatment and generally resolved within 1 day.¹⁴ In another study evaluating patient response of 1058 new patients after 4712 treatments by chiropractors in Norway, 55% of the patients reported at least 1 adverse event after a treatment.¹⁵ Adverse events included local discomfort (53%), headache (12%), tiredness (11%), and radiating discomfort (10%). A majority of the adverse events were reported within 4 hours of the treatment, and almost 75% had resolved within 24 hours.¹⁵ The most recently published randomized controlled trial (RCT) looking at adverse events in manual therapy evaluated chiropractic spinal manipulative therapy for migraineurs.¹⁶ For the 703 interventions applied to 70 participants, the most common adverse events were local tenderness, tiredness, and neck pain. The incidence of adverse events was 21% in the spinal manipulative therapy group and 8% in the placebo group; no serious adverse events were reported.¹⁶ The incidence rate of mild to moderate adverse events was approximately 41% in cohort studies and 22% in RCTs.¹⁰ The lower incidence rate in RCTs was attributed to underreporting because adverse events were not the primary outcome and were often poorly described.

To our knowledge, only 1 small study prospectively documented adverse events after osteopathic manipulative therapy (ie, manipulative care provided by foreign-trained osteopaths).¹¹ The evaluation of patient-reported “additional effects of treatment” 10 minutes after treatment from 63 new patients at a teaching clinic in the United Kingdom found that patients most commonly reported pain (52%) and stiffness (25%) in the area(s) treated, followed by unexpected tiredness (15%), lightheadedness (15%), radiating pain (13%), and headaches (8%).¹¹ The generalizability of the results was limited by the small sample size and the inexperience of the treatment providers, who were fourth-year osteopathy students.

Two of the largest RCTs assessing the effectiveness of OMT reported adverse events.^17,18 The OSTEOPATHIC Trial,¹⁷ evaluating the efficacy of OMT in treating participants with chronic low back pain, included 6 protocol-based OMT sessions during an 8-week intervention. Among all participants, there was a 6% incidence rate for adverse events and 2% for serious adverse events. None of the serious adverse events were adjudicated by an independent safety officer to be definitely or probably related to the OMT. In the OMT group, 3% of participants developed contraindications for continued participation, including 1 participant who experienced recurrent back spasticity after OMT. Overall, there were no differences in the rate of adverse events between the OMT and ultrasound therapy groups.¹⁷

Another RCT evaluated a standardized OMT protocol given twice per day as adjunctive care for hospitalized elderly patients with pneumonia.¹⁸ Of the 130 participants in the OMT group, 22% reported musculoskeletal soreness or pain vs 14% in the sham and control groups. Three serious adverse events occurred in the OMT group, but an independent data and safety monitoring board did not attribute them to OMT.¹⁸

No serious adverse events were identified in the current study. The incidence of minor or moderate adverse events was 2.5%. Several factors should be considered when comparing the results of the current study with previously reported incidence rates. The timing of adverse event reporting may influence incidence rates and must be considered when comparing our results with other studies. The OMT techniques currently used in practice were more diverse than the chiropractic high-velocity thrust techniques evaluated in previous studies.^10,19 Additionally, results from RCTs are not typically comparable to those achieved in clinical practice. In RCTs, the characteristics of the study sample and the treatments given are tightly controlled. However, the population of patients receiving OMT at typical clinics is more diverse. Additionally, a common view within the osteopathic medical profession is that standardized OMT protocols do not represent the true osteopathic approach to patient care, in which treatments are defined by patient-specific physical findings identified during each office visit. Protocol-mandated OMT could require unnecessary or inappropriate treatment of some body regions, thus causing new or worsening of symptoms, which could result in a higher number of adverse events in the RCTs compared with the current study's findings.

In contrast to commonly held beliefs, the current study found adverse events were less likely to occur after office visits during which direct thrust HVLA techniques were performed than office visits during which HVLA was not performed. Conversely, when more subtle techniques like balanced ligamentous tension/ligamentous articular strain, indirect/functional, and cranial techniques were performed, the chance of having an adverse event increased. Although it is common to perform multiple OMT techniques within a single office visit, the sample size in the current study was not large enough to allow us to examine the influence of more than 1 technique at a time. Previous studies in which more homogeneous sets of techniques were used did not encounter this issue.^12,20-31 Other factors to consider when interpreting these data are the impact of the patient's health status and diagnosis on the choice of techniques used. For example, patients treated with HVLA may have been healthier, making them less prone to experiencing an adverse event, while balanced ligamentous tension/ligamentous articular strain techniques may have intentionally been chosen for more frail patients. A more extensive examination of the chief complaints and diagnoses related to choice of OMT techniques is necessary to determine the influence of these factors on the incidence of adverse events.

Strengths of the current study are that it was prospective and included data from a large group of clinicians, more than 1800 “real world” office visits, and a broad age range of adult patients. Given the definition of an adverse event as temporally related to treatment but not necessarily caused by it,⁶ and considering that many factors can influence posttreatment experiences, determining causal association between OMT and adverse events is challenging. To demonstrate this challenge, a retrospective review of more than 23 million elderly patients concluded that the risk of organic injury temporally associated with chiropractic spinal manipulation within 1 week after treatment was 76% lower than within 1 week after primary care.³² By focusing on adverse events reported immediately after treatment, we minimized possible confounding variables and maximized the likelihood of identifying an adverse event caused by OMT.

The current study had several limitations. Current evidence indicates that the overall incidence of serious adverse events in manual therapies is quite low.^10,19 Based on an estimated rate of 1.3 serious adverse events per 1000 manual therapy sessions,¹⁰ at least 4841 such sessions would need to be studied to have a 95% chance of observing at least 3 serious adverse events. In light of that estimate, the current study's sample size was too small to make a definitive determination of the incidence of serious adverse events immediately after OMT.

Another limitation is the number of incomplete data sets. For 44% of the office visits where an adverse event was reported, patients did not specify what type of symptoms they were experiencing. Consequently, the incidence rates of the types of adverse events reported were likely underestimated. Future studies need to be designed to ensure patients provide complete information to thoroughly assess the type and severity of adverse events.

The current study's participants were racially homogeneous (92% white) and likely consisted of established patients, which limits the generalizability of our outcomes. Interobserver reliability for diagnosis of somatic dysfunction was not assessed, and we assumed that the clinicians performed and documented OMT techniques in a similar manner. Assuring standardization of data collected must be continually addressed to ensure scientific rigor.

Future research evaluating safety of OMT and other forms of manual therapy will require researchers to face several challenges. First, the duration of effect from manual therapy during which adverse events are monitored needs to be defined to accurately assess adverse events and reactions. Second, it must be determined whether all adverse events are truly adverse. Clinician and patient experiences suggest that some level of adverse events may be necessary to achieve overall improvement. Following patients for some period after manual therapies is necessary to begin addressing these challenges.

Conclusion

The aim of medicine is to provide safe, effective, efficient, timely, patient-centered, and equitable care.³³ To achieve this aim in the diverse disciplines categorized under manual therapies, ongoing research is necessary to assess efficacy and safety for patients of varying ages and levels of health. Results of the current study contribute to a better understanding of adverse events associated with OMT. Sustaining research in this area is critical to help clinicians better inform their patients about what to expect after OMT, to inform the insurance industry and policy makers about the benefits and risks associated with OMT, and to achieve the overall aim of medicine.

Author Contributions

All authors provided substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; drafted the article or revised it critically for important intellectual content; gave final approval of the version of the article to be published; and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.