Why don’t more physicians use osteopathic manipulative medicine? A cross-sectional study of utilization and referral barriers

Survey development and distribution

The Mayo Clinic Institutional Review Board reviewed the study design and deemed it exempt (IRB#: 21–006292). The survey was developed by the authors, who are practicing academic osteopathic physicians, to assess attitudes toward OMM among osteopathic physicians, nonosteopathic physicians, and advanced practice providers (MDs, Bachelor of Medicine, Bachelor of Surgery [MBBSs], nurse practitioners [NPs], physician assistants [PAs]). The survey aimed to determine the proportion of DOs who perform OMM as part of their routine practice, as well as referral patterns, willingness to refer, and knowledge of the referral processes for OMM services. The ultimate goal was to investigate factors contributing to limited patient access to osteopathic manipulation services. The final survey can be seen in Appendix 1.

Survey items were adapted from prior studies investigating similar topics, with modifications to enhance relevance to the study setting [12], [13], [14]. Because no previous surveys had established substantial validity evidence, the authors conducted an iterative refinement process. Pretesting was conducted with a diverse group of physicians and researchers within the same health system, including MDs, DOs, residents, and experts from the Mayo Clinic Survey Research Center. Participants provided feedback on item clarity, response burden, relevance, and comprehensiveness, leading to modifications in wording, pacing, and question targeting. Following pretesting, the Mayo Clinic Survey Research Center assisted in finalizing and distributing the survey. The survey was administered during the month of September 2022.

Statistical analysis

The primary aim of the analysis was descriptive, focusing on estimating the percentage of clinicians who feel that a particular factor (e.g., “time”) is a barrier to utilizing OMT, particularly among DOs. Based on previous research, we assumed that 40 % of DOs feel that time is a barrier to OMT for purposes of sample size calculation. With 50 DO respondents, we would be able to estimate this percentage with a +/− 14 percentage-point margin of error (MOE) based on a 95 % confidence interval (MOE=2 * sqrt [0.4 * 0.6/50]). Further, with 50 respondents, we would have 82 % power to detect a difference of +/− 20 percentage points from the observed percentage to a hypothesized percentage, based on a one-sample test for a proportion (with a two-sided type-I error rate of 5 %).

To achieve 50 respondents in each group, we planned to survey an equal number of non-DO respondents. Assuming a response rate of 30 % for DOs and 10 % for non-DOs, surveys were distributed to 952 individuals, including all 184 DOs in the region and a random sample of 768 non-DOs. Responses with missing data for a given question were excluded from the analysis of that specific item. The descriptive analysis was reported as raw percentages where applicable. Select Likert-style items were visualized utilizing stacked bar charts, ordered by the total number of positive responses in descending order.

Participant recruitment

Participants were recruited internally from the Mayo Clinic Midwest region, a large, multistate, healthcare system. Individuals met the inclusion criteria if they were actively employed by Mayo Clinic Rochester or Mayo Clinic Health System as medical providers, including DOs, MDs, MBBSs, NPs, and PAs. There were no separate exclusion criteria. Participants were contacted utilizing internal email, and were not compensated for participating in the study. The study was developed utilizing the Research Electronic Data Capture (REDCap) web-based application. All respondents provided written informed consent, and all responses were anonymized. No personal identifying information was collected.

Respondent demographics

Most respondents (n=115, 68.9 %) practiced outpatient medicine, and nearly half (n=72, 43.1 %) practiced inpatient medicine. The most common specialty among DO respondents was family medicine (n=21, 28.4 %), whereas the most common specialty for non-DO respondents was ‘internal medicine subspecialty’ (n=19, 21.3 %) (Table 1).

OMM usage among DOs

Among the 76 DO respondents, 64.0 % (n=48) graduated between 2001 and 2014 (Table 1). About a quarter of the respondents (n=21, 28.0 %) reported utilizing OMM. Of those who utilize OMM, the largest group (n=8, 38.1 %) reported utilizing it on 1–5% of their patients (Table 2), 52.4 % (n=11) expressed a desire to spend more time performing OMM, 66.7 % (n=14) were interested in educational activities to advance OMM skills, and 80 % (n=16) supported the use of OMM by osteopathic medical students or residents rotating with them (Table 2).

Perceived barriers to OMM

To assess barriers to utilizing OMM, 12 statements were posed to all DO respondents (Figure 1). Responses indicated that 94.6 % (n=71) disagreed or strongly disagreed with the statement “I had a lack of OMT training in medical school,” but 69.3 % (n=52) agreed or strongly agreed with “I had a lack of OMT training in residency.” Most respondents felt that OMT was relevant to their specialty and did not face difficulty obtaining privileging. Two of the most common reasons given for not utilizing OMT were insufficient time during patient encounters (n=51, 68.0 %) and competing professional interests (n=59, 79.7 %). Opinions on OMT reimbursement were ambivalent, with 64.8 % (n=48) neither agreeing nor disagreeing that “OMT offers poor reimbursement.”

Figure 1:

Responses from osteopathic physicians regarding barriers to the use of OMT in their practice (n=76).

Conditions treated with OMM

DOs who utilize OMM reported they would be somewhat likely or very likely to utilize it to treat headache (n=18, 85.7 %) and neck pain (n=19, 90.5 %), although they were unlikely to utilize it for abdominal pain (n=16, 76.2 %), shortness of breath/asthma/chronic obstructive pulmonary disease (COPD) (n=14, 66.7 %), and newborn feeding difficulties (n=13, 61.9 %) (Figure 2). This aligns with conditions that clinicians who do not utilize OMM are willing to refer for, with many reporting that they would be very likely or somewhat likely to refer for back pain (n=88, 63.3 %) and neck pain (n=78, 55.7 %) but not for constipation (n=14, 9.9 %), newborn feeding difficulties (n=14, 9.9 %), abdominal pain (n=11, 7.9 %), or shortness of breath/asthma/COPD (n=7, 5.0 %) (Figure 3).

Figure 2:

The likelihood that DOs who perform OMT would do so for the treatment of various clinical conditions (n=21).

Figure 3:

The likelihood of non-OMM providers to refer to OMM for the treatment of various clinical conditions (n=146).

Non-DO clinicians’ attitudes

Most non-DOs reported limited or no exposure to clinicians who do perform OMT (61, 67.0 %) (Table 2). Only 13.2 % (n=12) agreed or strongly agreed that they felt confident in their ability to refer patients to OMT for appropriate conditions. Only 17 (18.7 %) agreed or strongly agreed that they would allow osteopathic medical students to utilize OMT during rotations. The majority of non-DOs (n=44, 51.2 %) agreed or strongly agreed that they would be interested in attending educational activities that advanced their ability to refer appropriate patients for OMT.