Physical therapists’ attitudes are associated with their confidence in and the frequency with which they engage in prescription opioid medication misuse management practices with their patients. A cross–sectional study

Abstract

Background:

In the US, prescription opioid medication misuse (POMM) necessitates engagement of physical therapists (PTs). We (1) evaluated the attitudes of (PT) related to their management of patients with POMM and (2) examined the association between these attitudes and PTs confidence in POMM–related management abilities and the frequency with which they engaged in POMM–related management practices.

Methods:

We conducted a national survey of PTs that included a modified Drug and Drug Problems Perception Questionnaire (DDPPQ). Confidence in POMM–related abilities and the frequency of engaging in POMM–related management practices were measured. Logistic regression evaluated the association between the DDPPQ subscales (role adequacy, role legitimacy, role self–esteem, role support, job satisfaction) and confidence and frequency outcomes.

Results:

The analysis included 402 respondents. Role adequacy and legitimacy subscales were associated with confidence and frequency outcomes (p<.05), indicating that more favorable role adequacy and legitimacy attitudes are associated with greater odds of having more confidence in POMM–related management abilities and of engaging in more frequent POMM–related management practices.

Conclusions:

PTs with a greater sense of preparedness to engage in POMM–related management were more likely to report greater confidence in POMM–related management abilities and engage in POMM–related management practices with greater frequency.

Introduction

Misuse of prescription opioids includes using opioids other than prescribed or using opioids longer than intended^1,2 and opioid use disorder (OUD) or addiction occurs when the patient attempts unsuccessfully to control their opioid use and social problems and failure to fulfill obligations is a result.³ For the purposes of this paper, we collectively refer to opioid misuse and OUD as prescription opioid medication misuse (POMM). In the US, approximately 9.9 million individuals misused prescription opioids in 2018,⁴ and nearly 30% of patients taking long–term prescription opioid medications for painful orthopedic conditions may be misusing them.⁵ The diagnosis and treatment of opioid misuse has relied primarily on physicians being properly trained in opioid risk assessment, risk mitigation, and addiction.⁶ Recently, physician and nurse leaders in addiction medicine have advocated for other health care provider types such as nurses, public health professionals and counselors to be trained to diagnose substance use disorders and to use non–pharmacologic approaches to manage substance use.⁷ Physical therapists are members of the health care team that commonly manage patients with painful orthopedic conditions taking prescription opioid medications.⁸ Physical therapists, therefore, could play a prominent role in screening, diagnosing and treating patients with POMM.

Physical therapy is among the nonpharmacological alternatives recommended by the US Centers for Disease Control and Prevention for the management of chronic pain.⁹ The American Physical Therapy Association promotes physical therapists as providers who play a key role in responding to the US opioid crisis¹⁰ and physical therapy profession opinion leaders recognize that physical therapists can play a role in helping to solve the opioid epidemic in the US.^11,12 For example, in patients with spine pain, initial management in physical therapy may prevent some patients from initiating opioid therapy,^13,14 which may protect these patients from engaging in POMM that could lead to opioid diversion, overdose, and death.^1,15 Physical therapists manage patients with orthopedic conditions via referral from other providers and as first contact clinicians.^16-18 Up to 35% of patients are taking prescription opioids when they attend physical therapy for common conditions such as spine pain^8,19,20 and osteoarthritis.²¹ Some of these patients may be engaging in POMM while concurrently attending physical therapy.

Little is known about how physical therapists perceive their role in managing patients engaging in POMM or how their attitudes influence the confidence and frequency with which these physical therapists engage in POMM–related patient management practices. Therefore, the purposes of this study were to: (1) evaluate the attitudes of physical therapists related to their management of POMM in their patients with orthopedic conditions and (2) examine the association between these attitudes and physical therapists’ confidence in their POMM–related management abilities and the frequency with which physical therapists engage in POMM–related management practices.

Materials and methods

Survey construction

This research used a cross–sectional study design. Existing and modified instruments were used to survey members of the Academy of Orthopedic Physical Therapy (AOPT); a US organization that promotes the education, research and practice of orthopedic physical therapy.²² At the beginning of the survey, POMM was defined as: using prescription opioid medications without a prescription (i.e., the patient is using an old prescription or taking prescription opioid medications prescribed to someone else), or using prescription opioids for a reason other than the condition for which they were prescribed, or using prescription opioids at higher doses, more often, or for a longer period than prescribed, or being addicted to prescription opioid medications. The survey included: (1) clinician demographics, (2) clinician’s current practices related to the management of patients with orthopedic conditions who may be engaging in POMM, (3) clinician’s confidence in managing patients with orthopedic conditions who may be engaging in POMM, (4) the frequency of management practices related to patients with orthopedic conditions who may be engaging in POMM, and (5) physical therapists’ attitudes regarding the management of patients engaging in POMM.

Outcome variables

The frequency of 5 POMM–related management practice questions were our outcome variables. These questions were adapted from a survey developed to improve health care providers’ management of patients with alcohol problems.^23,24 For example, a question related to the frequency with which an health care provider asked about alcohol use was: “How often do you ask patients about alcohol problems?”²⁴ This question was modified to: “How often do you ask patients about problems with prescription opioid medication misuse? Four additional frequency of management practice questions were asked and similarly modified ([1] How often do you formally screen patients for prescription opioid medication misuse using screening instruments, such as the Current Opioid Misuse Measure or other tools; [2] How often do you assess patients’ readiness to change their prescription opioid medication misuse behaviors?; [3] How often do you discuss/advise patients to change their prescription opioid medication misuse behaviors?; [4] How often do you refer patients with prescription opioid medication misuse for further assessments or interventions?). The frequency outcome variables included the following 7 Likert scale response options: Never, Rarely (less than 10% of the time), Occasionally (about 30% of the time), Sometimes (about 50% of the time), Frequently (about 70% of the time), Usually (about 90% of the time) or Every time.

Additional outcomes were the physical therapists’ confidence in 5 POMM management–related abilities which were similarly adapted:²³ ([1] I am confident in my ability to ask patients about prescription opioid medication misuse, [2] I am confident in my ability to formally screen patients for prescription opioid medication misuse using screening instruments, such as the Current Opioid Misuse Measure or other tools; [3] I am confident in my ability to assess patients’ readiness to change their prescription opioid medication misuse behaviors; [4] I am confident in my ability to discuss/advise patients to change their prescription opioid medication misuse behaviors; [5] I am confident in my ability to refer patients with prescription opioid medication misuse for further assessments or interventions?). Each confidence with management practice question was scored on a 7–point Likert scale (1 = no confidence; 7 = high confidence).

Independent variables

The revised Drug and Drug Problems Perception Questionnaire (DDPPQ) is a 20 item valid and reliable tool used to measure the attitudes of health care providers related to the management of patients misusing drugs and contains 5 subscales.²⁵ The “Role Adequacy” (7 items) subscale assesses whether the respondent felt they were adequately prepared to work with patients misusing drugs. “Role Legitimacy” (2 items), assesses respondents’ attitudes about whether they have a right to obtain drug use information from patients. “Role Support” (3 items) focuses on attitudes about the working environment for managing patients misusing drugs. “Role Self–esteem “(4 items) measures the respondent’s feelings of purpose in working with patients misusing drugs and “Job Satisfaction” (4 items) measures the respondent’s sense of satisfaction they from working with patients who misuse drugs.^9,25 These DDPPQ subscales were our predictor variables of interest. Each item is scored on a 7–point scale (1 = strongly agree to 7 = strongly disagree). Higher scores for each subscale indicate more negative attitudes of working with patients who misusing drugs. Each DDPPQ question was adapted to reflect the POMM focus of this study. For example, a DDPPQ question reads “I feel I have a working knowledge of drugs and drug related problems” and was modified to “I feel I have a working knowledge of prescription opioid medication misuse and its related problems.”

Covariates

We chose covariates for their hypothesized association with the outcome variables. Demographic variables included sex and years since graduation from first professional physical therapy school. The percentage of the respondents’ clinical practice that was comprised of orthopedic patients was measured in 20%–point increments (0–20%, 21–40%, 41–60%, 61–80%, 81–100%), the percentage of patients taking prescription opioid medication for pain who were managed by the respondent was measured in the same 20%–point increments. The perception that physical therapists can play a role in modifying POMM in their patients was measured on a 7–point Likert scale from strongly disagree to strongly agree then dichotomized for the analyses (0 = strongly disagree – neither agree or disagree, 1 = somewhat agree – strongly agree). The hours of opioid misuse training that respondents received was categorized (0 hours, 1–6 hours, 7–25 hours, 26–90 hours, > 90 hours) then dichotomized for analysis (0 = no hours of training; 1 = any hours of training). Specialty certification was categorized as whether the respondent held an orthopedic certified specialist (OCS) or was a Fellow of the American Academy of Orthopedic Physical Therapists (FAAOMPT). Physical therapists achieved the OCS through completion of an accredited orthopedic physical therapy residency program or through maintaining a minimum number of hours managing patients with orthopedic conditions and passing a structured exam. FAAOMPTs successfully completed an accredited post–professional orthopedic and manual physical therapy fellowship program.

Survey distribution

Eligible respondents practiced physical therapy in the US and were AOPT members. The survey was targeted toward members of the AOPT members were targeted for the survey based on the likelihood that they manage patients with orthopedic conditions who were take prescription opioid medications for pain.

We used the AOPT email listserv to distribute the survey. One week after the survey invitation, the AOPT posted reminder invitations to its members using social media (e.g., Twitter, Facebook and Instagram). Respondents were instructed not to complete the survey more than once. As an incentive, potential participants were alerted that for every 5 surveys completed, $1.00 would be donated to the Foundation for Physical Therapy.

Interested participants clicked on an electronic link connecting them to an informed consent document. After indicating consent, the participant was able to complete the survey. Responses were collected over a 3–week period starting June 18, 2020. The University of Utah’s Institutional Review Board approved this study, which was determined to be exempt under category 2. The Checklist for Improving the Quality of Web Surveys was used in reporting the development, distribution and results of the survey.²⁶

Data analysis

Descriptive statistics were used to characterize the sample. Separate logistic regression models were constructed to examine the association between the each DDPPQ subscales and each frequency and confidence outcome. For our main analysis, we were interested in whether the DDPPQ subscales were associated with a low versus high frequency of engaging in POMM–related management practices and low versus high confidence with these management practices. Each of the 7–response options for all frequency outcomes ranged from Never to Every time. Therefore, we conceptualized “low frequency” as representing the 1st (Never) through the 4th (Sometimes) response options and “high frequency” as representing the 5th (Frequently) through the 7th (Every time) response options. Similarly, each of the confidence outcomes had response options ranging from 1–7. We conceptualized “low confidence” as any rating of 1 through 4 and “high confidence” as any rating of a 5 through 7. Each frequency and confidence outcome was then dichotomized into low (0) versus high (1). The goodness–of–fit (GoF) test assessed whether the data fit each logistic regression model and residual plots were used to visualize the data and assess for outliers. If the GoF test was significant (p<.05) outliers were removed and the model was run again and GoF reassessed. Regression results are expressed in adjusted odds ratios with 95% confidence intervals after controlling for the other DDPPQ subscales and covariates (years since graduation from first professional physical therapy school, sex, OCS or FAAOMPT, percent of orthopedic patients in clinical practice, percent of patients in clinical practice who are taking prescription opioids, perception that physical therapists can play a role in managing POMM and hours of opioid misuse training).

Some of the frequency and confidence responses were clustered at the low end of the scales.

Because our results may differ based on how we dichotomized the frequency and confidence outcome variables, we reran alternative logistic regression models with each frequency outcome dichotomized into low frequency (scores ranging from Never through Rarely) and high frequency (scores ranging from Occasionally through Every time) and each confidence outcome was dichotomized into low confidence (scores ranging from 1 to 2) and high confidence (scores ranging from 3 to 7).

Results

The AOPT email invitation was sent to 17,898 physical therapists with 402 included in the final analysis. The reasons for exclusion are included in the enrollment Figure 1. Each logistic regression model presented has a GoF test of p≥.05 after outliers were removed.

Figure 1.

Reasons for exclusion from analysis

Respondents profile

The respondents’ demographics are presented in Table 1. All states and the District of Columbia were represented with the exception of AK, HI, ME, RI, SD and VT. The mean years of experience was 15.3 (SD = 11.7) and 58.1% were either an OCS or FAAOMPT. Over 60% of the respondents managed a caseload of orthopedic patients for which they estimated up to 40% took prescription opioid medications for pain.

Table 1.

Respondent Demographics.

Characteristic	n (%)
Age, y, (sd)	41.0 (11.2)
Gender (female)	203 (50.4%)
Years since graduation, y, (sd)	15.3 (11.7)
Race and Ethnicity
American Indian or Alaska Native	1 (0.24%)
Asian	11 (2.73%)
Black or African American	2 (0.5%)
Hispanic or Latino	13 (3.2%)
Hawaiian or Other Pacific Islander	2 (0.5%)
White	378 (93.8%)
Region of practice
West (AK, AZ, CA, CO, HI, ID, MT, NM, NV, OR, UT, WA)	141 (34.0%)
South (AL, AR, DE, FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, )	97 (24.1%)
Midwest (IL, IN, IA, KS, MI, MN, MO, NE, ND, OH, SD, WI)	81 (20.1%)
Northeast (DC, CT, MA, ME, NH, NJ, NY, PA, RI, VT)	74 (18.4%)
Certification
AAOMPT Fellow or Orthopedic Certified Specialist	234 (58.1%)
Percent MSK practice
0—20%	57 (14.2%)
21—40%	20 (5.0%)
41—60%	20 (5.0%)
61—80%	55 (13.7%)
81—100%	250 (62.2%)
Percent of MSK patients prescribed opioid medication
0—20%	104 (25.9%)
21—40%	141 (35.1%)
41—60%	115 (28.6%)
61—80%	33 (8.2%)
81—100%	6 (1.5%)
How often do you ask your patients if they have a history of prescription opioid medication misuse? For example, "Have you ever used more of the opioid medication that prescribed?" Or "Have you ever had to go to more than 1 provider to fill your opioid prescription?
Never	144 (36.0%)
Rarely – less than 10% of the time	135 (34.0%)
Occasionally – about 30% of the time	46 (11.4%)
Sometimes – about 50% of the time	32 (8.0%)
Frequently – about 70% of the time	13 (3.2%)
Usually – about 90% of the time	13 (3.2%)
Every time	15 (3.7%)
Physical therapists can play a role in modifying prescription opioid use in patients who are misusing prescription opioids.
Strongly agree	150 (37.3%)
Agree	142 (35.2%)
Somewhat agree	66 (16.4%)
Neither agree or disagree	17 (4.2%)
Somewhat disagree	10 (2.5%)
Disagree	3 (0.7%)
Strongly disagree	11 (2.7%)
Hours of opioid misuse training
0	114 (28.4%)
1—6	204 (50.8%)
7—25	64 (15.9%)
26—90	10 (2.5%)
More than 90hours	7 (1.7%)

Abbreviations: AAOMPT: American Academy of Orthopedic and Manual Physical Therapists.

Missing in “Race”=9.

Missing in “Region”=9.

Missing in “Percent of MSK patients prescribed opioid medication”=3.

Missing in “How often do you ask your patients if they have a history of prescription opioid medication misuse”=4.

Missing in “Physical therapists can play a role in modifying prescription opioid use in patients who are misusing prescription opioids.”=3.

Missing in “Hours of opioid misuse training”=3.

For the DDPPQ, the mean score for each subscale was below the midpoint in the possible range of scores for each subscale indicating the responses trended toward more favorable attitudes (Table 2).

Table 2.

DDPPQ Total and DDPPQ Subscale Scores.

DDPPQ	Possible range	Mean	SD	Range
Total	20—140	60.8	(17.7)	20—130
Role adequacy	7—49	23.6	(8.9)	7—48
Role legitimacy	2—14	4.4	(2.2)	2—13
Role support	3—21	9.8	(4.6)	3—21
Role self–esteem	4—28	11.0	(4.0)	4—25
Job satisfaction	4—28	12.3	(3.9)	4—28

Abbrevations: DDPPQ: Drug and Drug Problems Perception Questionnaire; SD: standard deviation.

Confidence and frequency

Respondents had the least confidence in screening for POMM (M = 3.1, SD = 1.8) and in assessing readiness to change POMM–related behaviors (M = 3.7, SD = 1.6). Respondents had the greatest confidence in asking patients about POMM (M = 4.7, SD = 1.5) followed by referring for further treatment (M = 4.6, SD = 1.7) and discussing/advising about POMM (M = 4.1, SD = 1.6).

Greater than 50% of respondents reported that they engaged in each POMM–related management practice about 30% or less than the times when they could have (Table 3). The POMM–related management practice that respondents reported that they engaged in “every time” with the greatest frequency was referring patients for further POMM–related treatment (19, 4.7%) and the practice that respondents reported “never” engaging in with the greatest frequency was formally screening patients for POMM (254, 63.2%) (Table 3).

Table 3.

Frequency of Respondent Engagement In POMM–related Management Practices in Patients with POMM^a.

Frequency	Ask n (%)	Screen n (%)	Assess n (%)	Advise/discuss n (%)	Refer n (%)
Never	33 (8.2)	254 (63.2)	86 (21.4)	34 (8.5)	71 (17.7)
Rarely – less than 10% of the time	113 (28.1)	52 (12.9)	100 (24.9)	108 (26.9)	112 (27.9)
Occasionally – about 30% of the time	68 (16.9)	16 (4.0)	41 (10.2)	59 (14.7)	49 (12.2)
Sometimes – about 50% of the time	59 (14.7)	11 (2.7)	54 (13.4)	60 (14.9)	57 (14.2)
Frequently – about 70% of the time	38 (9.5)	16 (4.0)	44 (11.0)	48 (11.9)	26 (6.5))
Usually – about 90% of the time	28 (7.0)	2 (0.5)	25 (6.2)	33 (8.2)	25 (6.2)
Every time	18 (4.5)	7 (1.7)	8 (2.0)	18 (4.5)	19 (4.7)

POMM: prescription opioid medication misuse.

^aMissing in ‘ask’ = 45; missing in ‘screen’ and ‘assess’= 44; missing in ‘advise/discuss’ = 42; missing in ‘refer’ = 43.

Alternative logistic regression models

The results of our alternative logistic regression models are displayed in Supplementary Appendices 1 and 2. Our results are supported by the alternative analysis. Among the 5 separate original models with ‘frequency’ and ‘confidence’ outcomes, none of the 5 DDPPQ subscales that were significantly associated with any outcome changed in the direction of the association and was significant in the alternative models.

Association between DDPPQ subscales and frequency

Each DDPPQ aOR is interpreted as the change in odds of engaging in POMM–related management practices with higher frequency as attitudes increase by 1 unit. A 1–unit increase in the DDPPQ subscales means a more negative change in attitudes. Thus, an aOR that is less than 1 indicates increased frequency of management practices. The Adequacy subscale was significantly associated with each frequency outcome (p<.05) with the exception of “refer”. This result indicates that for every 1 unit increase in the adequacy subscale, respondents had lower odds of performing 4 of the 5 POMM–related management practices with high frequency. The Legitimacy subscale was associated with “ask” about (aOR = 0.72, 95%CI: 0.59–0.90, p=.003), “assess” for (aOR = 0.61, 95%CI: 0.49–0.78, p= <.001) and advise about (aOR = 0.64, 95%CI: 0.52–0.78, p= <.001) POMM. The Legitimacy subscale was not associated with any other frequency outcome. The Support subscale was associated only with “refer” for POMM (aOR = 0.90, 95%CI: 0.82–0.96, p=.011). The Self–esteem subscale was associated only with “assess” (aOR = 0.90, 95%CI: 0.81–0.99, p=.031) and with “advise” (aOR = 0.91, 95%CI: 0.83–0.96, p=.03) and the Job Satisfaction subscale was not associated only with any frequency outcome (Table 4).

Table 4.

Variables Associated With Frequency of Physical Therapists Engaging in POMM^a–Related Management Practices.

Variable	Frequency of ask about POMM aOR (95% CI) n=338	p	Frequency of screen for POMM aOR (95% CI) n=339	p	Frequency of assess for POMM aOR (95% CI) n=338	p	Frequency of advise about POMM aOR (95% CI) n=339	p	Frequency of refer for POMM aOR (95% CI) n=340	p
DDPPQb
Adequacy	0.90 (0.86—0.96)	<.001*	0.87 (0.79—0.97	.01*	0.94 (0.88—0.98)	.018*	0.90 (0.86—0.95)	<.001*	0.97 (0.92—1.02)	.21
Legitimacy	0.72 (0.59—0.90)	.003*	0.92 (0.65—1.29)	.63	0.62 (0.49—0.78)	<.001*	0.64 (0.52—0.78)	<.001*	0.92 (0.76—1.12)	.41
Support	1.00 (0.93—1.09)	.86	0.92 (0.80—1.06)	.25	0.98 (0.91—1.06)	.63	1.03 (0.96—1.110	.39	0.90 (0.82—0.96)	.011*
Self–esteem	0.92 (0.84—1.01)	.08	0.89 (0.75—1.06)	.17	0.90 (0.81—0.99)	.039*	0.91 (0.83—0.96)	.03*	0.91 (0.83—1.00)	.05
Job satisfaction	0.95 (0.87—1.05)	.34	0.96 (0.82—1.12)	.63	0.96 (0.87—1.95)	.34	0.98 (0.89—1.61)	.58	0.92 (0.84—1.01)	.07
Years since graduation	1.01 (0.99—1.04)	.20	1.02 (0.98—1.06)	.26	1.00 (0.97—1.03)	.45	1.02 (0.99—1.04)	.16	1.03 (1.00—1.05)	.047*
Sex (female)	1.06 (0.57—1.98)	.84	1.10 (0.39—3.13)	.84	1.29 (0.67—2.47)	.96	0.88 (0.48—1.61)	.67	0.81 (0.43—1.54)	.52
OCSc or Fellow	1.35 (0.70—2.58)	.36	1.14 (0.40—3.26)	.81	1.57 (0.79—3.10)	.19	0.71 (0.38—1.32)	.27	1.19 (0.62—2.32)	.60
Percent orthopedic patients	0.83 (0.69—1.01)	.07	0.90 (0.67—1.23)	.52	1.03 (0.83—1.27)	.79	1.14 (0.93—1.39)	.21	0.96 (0.79—1.16)	.65
Percent patients taking prescription opioids	0.95 (0.69—1.32)	.77	1.20 (0.73—1.93)	.46	0.95 (0.68—1.33)	.77	0.89 (0.65—1.23)	.49	0.70 (0.49—0.97)	.034*
Perception of role in management of POMM	0.58 (0.18—1.85)	.36	2.28 (0.21—24.59)	.49	1.95 (0.42—9.02)	.39	1.19 (0.34—4.13)	.78	0.53 (0.18—1.58)	.25
Hours of POMM training	1.91 (0.83—4.36)	.12	0.70 (0.19—2.26)	.60	1.45 0.63—3.30)	.38	1.70 (0.78—3.71)	.18	1.11 (0.51—2.44)	.78

^*Indicates significant values (p<.05); aOR: adjusted odds ratio; CI: confidence interval.

^aPrescription opioid medication misuse.

^bDrug and Drug Problems Perception Questionnaire.

^cOrthopedic Certified Specialist.

Association between DDPPQ subscales and confidence

For the associations between the DDPPQ subscales and the confidence outcomes, a 1–unit increase means a more negative change in attitudes. Thus, an aOR that is less than 1 indicates increased confidence in management practices. The Adequacy subscale was significantly associated with each confidence outcome (p<.001). This result indicates that for every 1 unit increase in the adequacy subscale, compared to respondents with low confidence in performing POMM management practices, respondents had greater odds of having high confidence in POMM management practices. The Legitimacy subscale was associated with “ask” about (aOR = 0.68, 95%CI: 0.57–0.81, p= <.001) and “advise” about (aOR = 0.64, 95%CI: 0.52–0.79, p= <.001) POMM but was not associated with any other confidence outcome. The Support subscale was associated only with “refer” for POMM (aOR = 0.87, 95%CI: 0.81–0.93, p= <.001). The Self–esteem subscale was associated only with “assess” (aOR = 0.86, 95%CI: 0.80–0.94, p= <.001) and the Job Satisfaction subscale was associated only with “refer” (aOR = 0.91, 95%CI: 0.83–0.91, p=.03) and “screen” (aOR = 0.88, 95%CI: 0.80–0.98) (Table 5).

Table 5.

Variables Associated With Confidence of Physical Therapists Engaging in POMM^a–Related Management Practices.

Variable	Confidence to ask about POMM aOR (95% CI) n=339	p	Confidence to screen for POMM aOR (95% CI) n=340	p	Confidence to assess for POMM aOR (95% CI) n=339	p	Confidence to advise about POMM aOR (95% CI) n=340	p	Confidence to refer for POMM aOR (95% CI) n=340	p
DDPPQb
Adequacy	0.91 (0.87—0.95)	<.001*	0.89 (0.85—0.95)	<.001*	0.85 (0.890—0.89)	<.001*	0.85 (0.81—0.98)	<.001*	0.92 (0.89—0.96)	<.001*
Legitimacy	0.68 (0.57—0.81)	<.001*	0.90 (0.74—1.11)	.32	0.92 (0.77—1.09)	.32	0.64 (0.52—0.79)	<.001*	0.88 (0.75—1.02)	.09
Support	0.99 (0.93—1.06)	.86	0.93 (0.85—1.01)	.07	0.98 (0.91—1.06)	.61	0.96 (0.09—1.04)	.33	0.87 (0.81—0.93)	<.001*
Self–esteem	0.86 (0.80—0.94)	.001*	1.07 (0.97—1.17)	.18	0.95 (0.87—1.05)	.34	0.94 (0.86—1.03)	.16	0.99 (0.91—1.07)	.72
Job satisfaction	0.94 (0.85—1.02)	.15	0.88 (0.80—0.98)	.014*	0.91 (0.82—1.02)	.12	0.96 (0.88—1.05)	.39	0.91 (0.83—0.99)	.032*
Years since graduation	1.02 (0.99—1.04)	.24	1.02 (0.40—1.44)	.21	0.98 (0.95—1.01)	.22	1.26 (0.68—2.33)	.46	1.24 (0.70—2.18)	.45
Sex (female)	0.54 (0.31—0.98)	.041*	0.76 (0.99—1.04)	.39	1.46 (0.75—2.86)	.25	1.02 (0.99—1.05)	.12	1.04 (1.02—1.07)	.001*
OCSc or Fellow	0.81 (0.45—1.47)	.49	1.04 (0.53—2.02)	.91	1.35 (0.69—2.64)	.38	0.9 (0.51—1.80)	.90	1.05 (0.59 to 1.870	.85
Percent orthopedic patients	1.29 (1.06—1.58)	.011*	0.85 (0.70—1.04)	.12	1.06 (0.84—1.34)	.60	1.05 (0.86—1.29)	.64	0.94 (0.78—1.44)	.55
Percent patients taking prescription opioids	0.98 (0.72—1.34)	.91	1.21 (0.87—1.70)	.25	0.76 (0.53—1.08)	.12	0.92 (0.66—1.27)	.62	0.99 (0.74—1.32)	.93
Perception of role in management of POMM	0.54 (0.20 t0 1.43)	.21	0.43 0.14—1.29)	.13	0.60 (0.20—1.77)	.35	0.65 (0.20—2.13)	.47	0.26 (0.10—0.71)	.008*
Hours of POMM training	0.90 (0.46—1.78)	.76	1.08 (0.48—2.41)	.85	1.15 (0.56—2.40)	.70	1.32 (0.62—2.88)	.46	1.54 (0.82—2.91)	.18

^*Indicates significant values (p<.05); aOR: adjusted odds ratio; CI: confidence interval.

^aPrescription opioid medication misuse.

^bDrug and Drug Problems Perception Questionnaire.

^cOrthopedic Certified Specialist.

Discussion

Given that physical therapists are well–positioned to manage POMM–related problems in their patients with orthopedic conditions, we sought to examine (1) the attitudes of physical therapists related to their management of POMM in their patients and (2) examine the association between these attitudes and physical therapist’s confidence in their POMM–related management abilities and the frequency with which physical therapists engage in POMM–related management practices. Because we defined POMM for the respondents as including opioid misuse and OUD, we have defined it similarly for the readers of this research. A notable finding from our work is that while the great majority of respondents (88.9%) “somewhat agreed”, “agreed” or “strongly agreed” that physical therapists can play a role in modifying POMM in their patients, 70.0% “rarely” or “never” ask patients if they have a history of POMM. This finding may illustrate that physical therapists are aware of the current opioid crisis in the US and see themselves as members of the health care team that could favorably influence POMM in their patients; however, they might feel that they lack specific skills to address POMM.

We found that the mean total DDPPQ score for the respondents was 60.8 (SD = 17.7), which indicates slightly favorable overall attitudes. However, there is room to improve the overall favorability of their attitudes. The mean role adequacy subscale score was 23.6 (SD = 8.9) with a possible range of scores between 7 and 49; and the mean role legitimacy subscale score was 4.4 (SD = 2.2) with a possible range of 2 to 14 (Table 2). When considering the mean item scores for the role adequacy and role legitimacy subscales, it appears that respondents may have more favorable views that their role in managing POMM in their patients is legitimate than their views about how adequately prepared they feel in such a role.

With training, physical therapist POMM–related patient management attitudes may change favorably, which could lead to better care of patients with POMM. The Commission on Accrediting Physical Therapy Education (CAPTE) provides no specific standards related to physical therapists screening or managing patients with POMM.²⁷ Therefore, it is unknown whether student physical therapists routinely receive such education. The rate at which licensed physical therapists seek POMM–specific training is also unknown. Training physical therapists could include educating them on the differences between opioid use, opioid misuse and opioid use disorder (OUD) and how to screen for opioid misuse and OUD. Physical therapists could be trained to provide a brief cognitive–behavioral intervention aimed at changing the patients’ POMM–related behaviors and to refer the patient to their physician or to an addiction specialist. Other non–physician healthcare providers have had their attitudes regarding managing patients with alcohol misuse and SUDs favorably altered after training. For example, after attending an 8–week graduate school course in how to recognize and treat SUDs, pharmacy students mean post–training DDPPQ scores for each subscale significantly improved (p<.05) compared to pretraining DDPPQ scores.²⁸ When non–physician healthcare graduate students, including physical therapy students, attended a 6–hour training course in screening, brief intervention and referral (SBIRT) for alcohol misuse and SUDs, post–training scores for the role adequacy, role legitimacy, role support and satisfaction subscales each significantly improved over pre–training scores, with the largest effect size for role adequacy (Cohen’s d = 1.06).²⁹ When nurses attended an 8–hour workshop on nursing practices for patients with SUDs, the role adequacy subscale demonstrated the greatest change from pre– to post–test (Δ= −14.7).⁹ Our results are unique in that we believe this study is the first to evaluate the attitudes that physical therapists have surrounding managing patients with POMM.

We found that decreases on some of the DDPPQ subscales (attitudes became increasingly more favorable) were positively associated with the respondents’ confidence in several of their POMM–related patient management abilities and their likelihood of engaging in some POMM–related management practices. As the attitudes of the respondents’ Role Adequacy improved, they were more likely to be confident in asking patients about POMM, formally screening patients for POMM, assessing the patients’ readiness to change POMM–related behaviors, advising about POMM and referring for further POMM–related treatment. These results are similar to research that found when rehabilitation counselors’ had favorable attitudes toward managing patients with SUD, they were more likely to have greater confidence in asking patients about substance use problems, screening for substance use and referring patients for further substance use treatment.²⁴ In our study, more favorable attitudes of legitimacy were associated with more confidence in asking patients about POMM and in discussing/advising patients about POMM but not the other “confidence” variables. This finding may reflect that respondents’ confidence in formally screening, assessing and referring about POMM is not influenced by their feelings that their role in managing POMM is legitimate.

The Role Support subscale was associated with the “refer” outcomes for both “confidence” and “frequency” but it was not associated with any other POMM outcomes. This result could reflect that some respondents worked in environments where addiction medicine specialists were available to receive referrals from physical therapists who identify patients in need of further care; or that they had colleagues, such as primary care physicians, that would provide further management. This interdisciplinary collaborative atmosphere could help to instill the respondents with a feeling of confidence to make referrals for POMM–related patient problems and the environment might be conducive to making these referrals with high frequency. It is possible that the respondents in our study did not need to feel support from their work environment in order to feel more confident in or engage in the other POMM–related management practices. When clinicians work in an interdisciplinary manner, provider confidence in the management of patients with OUD may improve. After integration of behavioral health into primary care, the confidence of primary care providers to deliver primary care based management of patients with OUD may improve.³⁰

We speculate that many physical therapists don’t function within health care environments that have access to addiction medicine providers or, if the physical therapists do have access, they do not practice in an interdisciplinary fashion that includes addiction services. Physical therapists who manage patients engaging in POMM should make efforts to establish clinical pathways that include providers of addiction medicine if referral is warranted.

This study has important limitations. Greater than 98% of the respondents were white, which may limit the generalizability of the results to physical therapists of other races. Respondents may have felt compelled to respond in a particular way to some survey questions even though their actual attitudes and practices may be different from how they responded. Therefore, social desirability response bias could have led to inaccurate self–reports.³¹ We did not assess practice heterogeneity. Some respondents might work in physical therapy practices that function in an interdisciplinary manner with providers of addiction medicine services while others do not, this potential practice heterogeneity among respondents could have also impacted our results. The cross–sectional nature of the study does not allow us to infer causality between the DDPPQ subscale scores and any “confidence” or “frequency” outcome. We targeted members of the American Physical Therapy Association’s AOPT; therefore, our sample may underrepresent all licensed physical therapists who encounter patients with POMM in the orthopedic setting, which could limit the study’s generalizability. We did not measure the respondents’ understanding of our definition of POMM, therefore, we cannot be certain that their responses actually reflect their knowledge of POMM. We are unaware whether the measures we adapted for our outcomes have been validated for POMM. Finally, although there were 402 respondents with eligible surveys for analysis, between 338 (83.9%) and 340 (84.4%) were included in the final regression models due to missing data and outliers.

Conclusion

Physical therapists are well–positioned to play a role in the management of patients engaging in POMM. As physical therapists’ view some aspects of their role in management of these patients more favorably, they have greater confidence in some POMM–related patient management abilities and they may engage in these abilities with greater frequency. Specialized POMM training could alter physical therapists’ attitudes of this role and increase their confidence and the frequency with which they engage in these management practices. Future research should focus on developing and implementing POMM training programs for physical therapists.