Investigating the safety and feasibility of osteopathic medicine in the pediatric oncology outpatient setting

Intervention

The total study duration consisted of 10 total weeks. Following enrollment, patients were approached to receive OMT at scheduled clinic visits for eight consecutive weeks. Osteopathic intervention lasted a maximum time allotment of 20 min per visit and was performed by a single osteopathic physician to ensure uniformity across visits and patients. A research team member administered validated constipation and pain scales. Participants were given an encrypted Apple iPad and completed pre- and postintervention surveys without research team assistance. These surveys consisted of the FACES pain scale and detailed questions about their bowel habits including the Bristol Stool Scale. The Bristol Stool Scale is a validated constipation scale utilized by participants ≥6 years of age for self-reported symptoms or reported by caregivers for children <6 years of age and administered at each visit prior to OMT [18]. It consists of scores 1–2 indicating clinical constipation, 3–4 representing normal stool, and 5–6 describing loose stool taking no more than 1 min to complete. The FACES pain scale was administered prior to OMT intervention to assess pain before OMT, and again after OMT to assess posttreatment pain, taking 10–30 s to complete [19]. The FACES pain scale is validated for children aged ≥3 years of age for self-reporting, and caregiver reporting for children <3 years of age [19]. Concomitant medications were updated at each visit, including details of as-needed medication use in the past week. Following each intervention visit, the charge nurse and bedside nurse were asked to complete a brief questionnaire via REDCap regarding interruptions in workflow. If a patient was admitted to an inpatient service or intensive care unit, OMT was not offered that week to maintain true feasibility of an outpatient intervention. Appropriate antibiotics, blood products, anti-emetics, fluids, electrolytes, and general supportive care including massage and physical therapy were utilized as clinically indicated by their primary medical team.

Osteopathic techniques

OMT consisted of myofascial release, muscle energy, balanced ligamentous tension, and visceral manipulation. Due to the collateral ganglia in the abdomen influencing regional visceral dysfunction, visceral manipulation included a ventral abdominal release and inhibitor pressure directed at the celiac, superior, and inferior mesenteric ganglia. Lymphatic pump and high velocity, low amplitude (HVLA) techniques were excluded due to the small hypothetical risk of inducing metastasis by increasing lymphatic flow in patients with assumed microscopic disease, and to protect the bony skeleton in patients receiving steroids and presumed decreased bone density [20].

Study endpoints

For the purposes of this study, feasibility is defined as the ability to incorporate OMT intervention into regularly scheduled chemotherapy outpatient appointments. Oncology clinics are multidisciplinary and require coordination between many providers, nurses, and administrative staff. The primary feasibility endpoint was defined as the percentage of patients that were available and agreed to receive OMT at two of their three ambulatory visits over the total 10-week study period. There are no previously published data to guide the definition of feasibility, therefore two of their three visits were decided among the study team to represent feasibility based on the belief that completion of over half of the eligible visits: (1) would indicate a reasonable degree of patient interest in receiving treatment, (2) show evidence of the flexibility to schedule OMT sessions without disruptions to the regular clinic workflow, and (3) may warrant further research in the benefit of investing hospital staff and resources to implement OMT as a supportive-care option for patients into oncology clinics. Safety outcomes were assessed at each study visit according to the required adverse event reporting tool, Common Terminology Criteria for Adverse Events (CTCAE) criteria, as mandated by the National Cancer Institute (NCI) [21].

Sample size

With an a priori goal of 23 enrolled subjects, there was approximately 80% power to detect if the percentage of subjects who agree to two of their three offered OMT treatments would be significantly greater than 50% utilizing a one-sided exact binomial test with a 5% type I error and an effect size of 25% (i.e., H₀:50% of patients agree to at least two out of three OMT treatments vs H₁:75% of patients agree to at least two out of three OMT treatments).

Statistical considerations

The proportion of subjects who agreed to two of the three offered OMT treatments was summarized and presented with a corresponding exact binomial 95% confidence interval, and the primary hypothesis was tested utilizing a one-sided exact binomial test. Descriptive statistics were utilized for additional safety and feasibility analyses. To be included in the safety analysis, patients must have received at least one OMT intervention. The feasibility analysis included patients who were eligible for at least six clinic visits. Adverse effects were noted at each study visit. The survey tools utilized included the Bristol Stool Scale for constipation and the FACES pain scale. A nonvalidated questionnaire was utilized to assess clinic flow and nursing views to better understand the feasibility of incorporating OMT in the clinic. The length of clinic visits, frequency of adverse events, nurse assessments of clinical flow, and patient satisfaction scores were summarized descriptively. All p values are two-sided and those less than 0.05 were considered statistically significant. Statistical analyses were performed utilizing SAS software, version 9.4 (SAS Institute, Cary, NC) or the base R statistical package (R Foundation for Statistical Computing, Vienna, Austria).

Participant characteristics

A total of 30 patients were screened for eligibility, 24 patients met the eligibility criteria, and 23 total patients were enrolled into the study, with one patient declining enrollment due to possible transfer of care outside of the state. The majority of patients were female (n=13, 56.5%), with a median age of 12 years at enrollment (range, 2.7–20.8 years). Patients had a variety of cancer diagnoses including leukemia (n=9, 39.1%), lymphoma (n=7, 30.4%), sarcoma (n=6, 26.1%), and neuroblastoma (n=1, 4.3%). Patient characteristics are summarized in Table 1.

Primary outcome safety

Patients exhibited adverse events prior to enrollment secondary to chemotherapy treatment, such as deep vein thrombosis, leg pain, and constipation. None of these prior conditions were noted to be exacerbated while the patients were in the study. No patient developed a serious adverse event, per the NCI guideline criteria, that was attributed to OMT (Figure 1). Unrelated adverse events including Clostridium difficile (n=1), typhlitis (n=1), and fatigue and nausea (n=1) were observed and attributed to expected disease and/or chemotherapy side effects.

Figure 1:

Summary of study visit and patient safety data.

Primary outcome feasibility

Of 23 enrolled patients, two were not included in the feasibility and safety analysis due to not having at least six evaluable visits. One patient withdrew shortly after enrollment per parental request, and one patient became ineligible prior to the first visit due to alteration in therapy that no longer included vincristine. Thus, a total of 21 patients were eligible to complete at least six visits and were included in feasibility analysis. All (100%) patients agreed to OMT in at least two of their three available treatment weeks (Figure 1), which is significantly greater than the null hypothesis of 50% of patients (p<0.0001). Among the 21 patients included in feasibility analysis, three patients were unable to complete the final three study visits – one patient was taken off study due to discontinuation of chemotherapy and transitioned to palliative care, and two patients’ final study visits were impacted by the COVID-19 pandemic. The majority of the missed OMT intervention studies were secondary to either canceled clinic visits due to chemotherapy-induced myelosuppression, or the OMT physician was unavailable to see the patient due to other commitments.

OMT intervention lasted an average of 14.2 min (range, 7.2–19.2 min). No bedside or outpatient charge nurse had concern for delays with clinic workflow during the visits. No patient was detained past their expected visit time to complete study requirements.

Descriptive patient reported data

Among the 148 total Bristol Stool Scales collected, the median was 3 at both the first and final visits. Nearly all patients (17 patients, 37 individual visits) had at least one occurrence of self-reporting no constipation although their Bristol Stool Scale indicated the presence of constipation with score of 1 or 2. However, there were no instances in which patients self-reported constipation but the score was reported as a 3 or greater. Overall, there was disagreement in 25% (37/148) of self-reported constipation and Bristol Stool Scale scores. When known, patients reported that they had a stool within 24 h after the previous OMT intervention at nearly every visit. Median FACES scores pre- and post-OMT were both 0. In 51 study visits, the FACES pain score decreased after OMT compared to the reported pain score prior to receiving OMT. There were zero instances in which the FACES pain score increased following OMT.

Limitations

Despite promising findings, the authors acknowledge the limitations of the study. This is a small sample size, single-institutional study that may not be representative of the wider pediatric oncology population or other oncology clinics. Nonetheless, our study was able to include a wide range of patients in age and cancer diagnoses.

Another limitation was that the lead investigator was an osteopathic physician, which could lead to implicit bias from patients. To best negate this, all quantitative survey tools and question responses were blinded and analyzed at the end of the study. We recognize the many limitations to descriptive analyses of OMT effects on constipation and pain, including modification of the prescribed bowel regimen, other supportive care modalities including massage and physical therapy, home diet implementations, recall bias with post-OMT bowel movements, and most notably the fact that patients receive different chemotherapy regimens with various timing, dosing, and frequency of vincristine.