Pediatricians’ Attitudes Toward and Use of Terminology About Mothers With Opioid Use Disorder

Abstract

Objectives:

Stigmatizing experiences for mothers with opioid use disorder (OUD) may impede health care engagement. We sought to characterize attitudes and terminology use among pediatric primary care clinicians as a potential target for improvement.

Methods:

We conducted a cross-sectional survey of 1004 clinicians at 28 clinics affiliated with 7 US residency programs (April to June 2022). Survey questions focused on trust, blame, and support for mothers with OUD were adapted from prior studies of self-reported attitudes, and terminology was categorized as preferred versus nonpreferred based on national recommendations. Frequencies were tabulated, and a 2-stage process of factor analysis and k-means clustering was used to group respondents by attitudinal responses. Multivariable logistic regression evaluated the association between participant characteristics and attitudinal groups.

Results:

Of 272 respondents (response rate 27.1%), most were female, non-Hispanic White residents; the distribution of clinical roles was similar among nonrespondents. Use of nonpreferred terminology was infrequent, although over 69% reported usual or occasional use of “substance abuse.” Analyses supported clustering respondents across 3 main constructs of trust, blame, and support. About 27% were categorized as having a low level of blame, and about 38% had a high level of trust. Adjusting for covariates, high trust was associated with confidence in training to provide care for opioid-exposed children (adjusted odds ratio [AOR] 1.84, P = 0.04), and low blame was associated with education on OUD stigma (AOR 3.43, P = 0.001).

Conclusions:

Pediatrician attitudes reflecting mistrust and blame toward mothers with OUD are not uncommon but may be addressed through training and education.

As rates of opioid use disorder (OUD) and overdose have increased dramatically across the United States over the last 20 years, the number of pregnancies affected by OUD has quadrupled.^1,2 Almost 3% of pregnant and postpartum Medicaid enrollees in the United States have documented OUD.³ Children with intrauterine opioid exposure (IOE) and mothers with OUD may face unique challenges, ranging from the transitional stresses of neonatal opioid withdrawal syndrome to chronic risks, such as poverty and maternal mental health conditions.^4,5 Pediatric primary care offers fundamental opportunities to provide longitudinal support for both child and parent, including screening for and addressing developmental, medical and social needs, providing anticipatory guidance, and supporting parenting skills and self-efficacy.⁶

However, stigma and perceived judgment from the health care team may serve as a barrier to prenatal and pediatric care engagement.^7–11 Negative attitudes toward persons with substance use disorder (SUD) are prevalent among medical professionals and have been found to decrease providers’ willingness to care for this population, to negatively affect the mental health of persons with SUD, and to deter engagement in SUD treatment.^12–15 In 1 study, pediatric trainees expressed blame for caregiver substance use and a lack of trust in mothers’ ability to appropriately engage in pediatric care while also supporting opportunities for recovery and the potential for successful parenting.¹⁶ A supportive parent-pediatrician relationship has been linked to behavior change regarding substance use and is among the most important aspects of well-child care from the perspective of children with IOE and parents with OUD.^17,18

A specific marker and potential mechanism of stigma and bias perpetuation within health care interactions is the use of stigmatizing language.¹⁹ Recognizing this, in 2022, the American Academy of Pediatrics (AAP) joined a chorus of other professional societies, releasing a policy statement that advocates for respectful communication that recognizes SUDs as chronic medical conditions through the use of “person-first” and “medically accurate terminology.”²⁰ However, little is known about the current use of preferred terminology among pediatric primary care providers or their attitudes toward parents with OUD.

This study aims to expand the understanding of stigma toward mothers with SUD among pediatric primary care clinicians by characterizing attitudes toward and use of terminology about mothers with OUD among a broad sample of pediatric practitioners and to explore potential factors associated with more positive attitudes.

METHODS

Study Population and Procedures

We conducted a cross-sectional survey of 1004 primary care pediatricians and pediatric trainees from 7 US children’s hospitals with pediatric residency programs. Network sampling was used to identify a diverse sample of programs with varying sizes, geographic regions, and practice settings. At each participating program, a local research team was responsible for study sample identification and recruitment. Eligible participants, including pediatric residents, attending pediatricians, and nurse practitioners in each program’s outpatient continuity clinic or primary care teaching sites, were recruited through an email containing a survey link and instructions. Both categorical and noncategorical pediatric residents were included. Reminder emails were sent biweekly to monthly by each team over a 3-month period from April to June 2022. The survey was conducted online using REDCap (Research Electronic Data Capture).^21,22 An electronic consent form preceding the survey emphasized that participation was anonymous and voluntary and provided contact information for the local study team. A small monetary incentive, in accordance with local institutional review board policies, was offered to survey participants. Personal identifiers were not collected within the survey; however, participants who opted to receive the monetary incentive were directed to a secondary survey link to collect contact information. These data were maintained in a separate database that was not linkable to their survey responses. All study procedures, including a waiver of documentation of informed consent, were approved by the respective IRB at each institution.

Survey Instrument

We developed a survey to assess pediatric clinician knowledge, attitudes, and practices for primary care of children with IOE and families affected by prenatal OUD. In total, the survey contained 48 items, including a combination of multiple-choice questions, Likert-like scale questions, and matrix questions. Local investigators at each of the 7 participating programs iteratively reviewed and revised the survey instrument. To ensure survey clarity and assess participant burden before dissemination, the instrument was pilot-tested with 3 primary care attendings and 3 residents at 1 site. Survey questions focused on clinician attitudes toward mothers with OUD and use of SUD-related terminology, as well as common diagnoses and issues faced by children with IOE, AAP-recommended screenings, and resources and referrals important to pediatric primary care of children with IOE.^23–25 Survey findings of clinician-reported screening practices, resources, and referrals are presented elsewhere.²⁶

Clinician Demographics

We assessed clinician characteristics (eg, demographics, prior OUD-related education, personal impact of SUD, and program affiliation) and practice-level characteristics (eg, payor mix and frequency of patients with IOE). The personal impact of SUD was assessed via a 5-point Likert scale (from “not at all” to “extremely”) using the question “To what extent have you, a family member, or close friend been diagnosed or personally impacted by a substance use disorder (including alcohol, prescription drugs, street drugs, and/or injection drugs)?”

Terminology Use

The survey asked about respondents’ use of SUD-related terminology. A list of 19 words, including both preferred and nonpreferred terms in random order, was developed based on study team consensus and the 2022 AAP policy statement.²⁰ Respondents were asked, “Which of the following terms do you use when referring to people with an opioid use disorder, either in personal or professional settings?” Answer options were “never,” “sometimes,” and “usually.” Terms were assigned into preferred or nonpreferred categories based on the AAP’s 2022 policy statement on terminology regarding substance use.²⁰

Attitudes

To assess attitudes toward mothers with OUD, respondents were prompted to indicate their level of agreement with 15 statements using a 5-point Likert scale, with answer options ranging from “strongly agree” to “strongly disagree.” Development of this portion of the survey was modeled after prior studies measuring respondent attitudes toward people with SUD,^16,27,28 with the tailoring of the statements for relevance to our study sample of pediatric clinicians and the topic of maternal OUD. Four statements, all beginning with the phrase “I am confident that…,” were derived from a previously published Trust scale measuring the confidence of pediatric trainees in mothers of infants with neonatal abstinence syndrome in the hospital setting.¹⁶ The statements were adapted to reflect a primary care setting. Five statements were taken from a previously published Punitive-Blaming scale, a series of statements focused on the culpability and parenting abilities of mothers with OUD.²⁹ This same previous study included a Supportive scale, with statements focused on treatment resources and recovery. From this scale, we utilized 6 statements most relevant to mothers with OUD in the primary care setting. Cronbach ɑ values for the internal reliability of each of these scales were previously reported as acceptable to good.^16,29 To minimize various forms of response bias, our survey presented the 15 statements in a mixed order. Given that these statements were adapted from previously developed scales for trust, blame, and support, we hypothesized that scoring the data based on these 3 same discrete constructs would be possible.

Statistical Analysis

We analyzed survey responses using descriptive statistics (ie, frequencies). The randomness of missing data was assessed based on the distribution of partial responses by survey item and by observable respondent characteristics. We then identified groupings of respondents based on patterns of responses to the 15 attitudinal statements using a 2-stage process. First, factor analysis was applied to the 15 statements to identify latent constructs explaining observed correlations among variables, using principal component analysis as the extraction method with varimax rotation. Next, a reduced set of component scores was used as input for the second stage of k-means clustering to group participants. An optimal number of clusters was determined by visual data assessment for clear separation and by plotting the reduction in variation after several runs of analysis with varying numbers of clusters. Analysis of variance compared variable means between clusters and scatterplots were then used to generate visual representations of the clusters. Lastly, we used multivariable logistic regression to evaluate the association between participant characteristics and cluster assignment. Statistical significance was set at a P value of < 0.05. To address missing data, we used a pairwise deletion method, omitting respondents from the analysis only for those questions which were left blank. Analyses were conducted in SPSS version 27.0 and STATA version 11.0.

RESULTS

Survey Participants

Surveys were distributed to 1004 clinicians from 28 primary care clinics affiliated with 7 pediatric residency programs located across 6 states. Residency programs ranged in size from 36 to 160 residents, with a median of 92 (interquartile range 64, 130). Of 329 surveys returned with either complete or partially complete responses (overall response rate 32.8%), 272 surveys (27.1%) were included in the final analytic sample based on the availability of data for at least 1 “terminology” or “attitude” question. Among these included 272 surveys, the percentage of missing data for each of the 15 attitude statements ranged from 0.4% to 1.8%, and for each of the terminology items, the percentage of missing data ranged from 0.4% to 1.5%. Excluded respondents who did not provide data for any of these survey items (n = 57) did not significantly differ by characteristics including age, clinical role, self-identified gender, institution, volume of patients with IOE, or personal impact of SUD (all χ² P values > 0.05).

As shown in Table 1, most of the analytic sample were residents (61.3%), 35 years and under (67.9%), identified as female (76.5%), White (75.4%), and non-Hispanic (93.7%). More than half of respondents reported that they had been personally impacted (at least slightly) by substance use among themselves, a family member, or a friend (62.0%). Most (80.5%) reported having received training on medications used to treat OUD, while 52.2% reported training on the impact of stigma, and 41.5% reported training on screening, brief intervention, and referral to treatment (ie, SBIRT). However, only 32.7% agreed that the training/education they had received was adequate to provide high-quality pediatric primary care to this population.

TABLE 1.

Respondent Characteristics

	% (n)
Clinical role (N = 271)
Attending physician	37.6 (102)
Nurse practitioner	1.1 (3)
Resident physician	61.3 (166)
Resident postgraduate year (PGY) level (N = 166)
PGY 1	30.1 (50)
PGY 2	33.7 (56)
PGY 3	34.3 (57)
PGY 4+	1.8 (3)
Resident training track (N = 166)
Categorical pediatrics	73.5 (122)
Medicine-pediatrics	9.0 (15)
Other noncategorical pediatrics	5.4 (9)
No response	12.0 (20)
Years post-residency/training (N = 105)
0–10 y	40.9 (43)
11–20 y	26.7 (28)
21–30 y	25.7 (27)
31 or more years	6.7 (7)
Age (N = 271)
35 y and under	67.9 (184)
36–45 y	14.0 (38)
46–55 y	10.0 (27)
56–65 y	6.6 (18)
66 y and older	1.5 (4)
Gender (N = 272)
Man	22.4 (61)
Woman	76.5 (208)
Gender non-conforming/Other/Prefer not to answer	1.2 (3)
Race (N = 272)
Asian/Native Hawaiian/Pacific Islander	18.0 (49)
Black/African American	4.0 (11)
White	75.4 (205)
Other	2.6 (7)
Ethnicity (N = 270)
Hispanic	6.3 (17)
Non-Hispanic	93.7 (253)
Personally impacted by substance use disorder (self, family or friend) (N = 271)
Not at all	38.0 (103)
Slightly	24.0 (65)
Somewhat	18.8 (51)
Very much	10.3 (28)
Extremely	8.9 (24)
Received education or training on: (N = 272)
Medications used to treat opioid use disorder	80.5 (219)
Screening, brief intervention, and referral to treatment (SBIRT)	41.5 (113)
Impact of stigma on people with opioid use disorder	52.2 (142)
Confident in adequacy of training (any agreement) (N = 269)	32.7 (88)
Practice type (N = 271)
Group practice/health maintenance organization	5.9 (16)
Hospital-owned practice/clinic	80.4 (218)
Non-profit community health center, public	11.1 (30)
Solo/2-physician practice	0.7 (2)
Unsure	1.9 (5)
Practice geography (N = 271)
Urban	78.2 (212)
Suburban	18.9 (51)
Semi-rural	3.0 (8)
Payor mix (N = 271)
Approximately equal mix of private and public	18.1 (49)
Mostly Medicaid insured	67.2 (182)
Mostly privately insured	10.0 (27)
Mostly uninsured	0.4 (1)
Unsure	4.4 (12)
Number of children with intrauterine opioid exposure seen by respondents in outpatient primary care in the last 12 mo (N = 272)
11 or more children	12.1 (33)
6–10 children	18.0 (49)
1–5 children	52.9 (144)
None	16.9 (46)

Assessment of Nonresponse Bias

Available information about the denominator of clinicians invited to participate demonstrates a similar distribution to respondents in the analytic sample with regards to clinical role (34.8% attendings/nurse practitioners, 65.2% residents; χ² comparison statistic 1.45, P value = 0.23) and residency training level (~210 residents in each of the PGY 1, 2, and 3 years, plus a small number of PGY 4 and 5; χ² statistic 0.43, P value = 0.81). However, we did observe differential responses by the institution (25.0%–84.8%), with an overrepresentation of respondents from the smallest program, which had the highest response rate (χ² statistic 20.75, P value = 0.002).

Use of Terminology

Rates of self-reported use of preferred and nonpreferred SUD-related terminology are shown in Figure 1. Results for 1 of 19 listed terms (“sobriety”) were omitted, as this term was subsequently deemed by the study team to be neither preferred nor nonpreferred based on the 2022 AAP Policy Statement. Use of nonpreferred terminology was infrequent, except for “relapse” (83.4% reporting use usually or sometimes), “substance abuse” (69.1% usually or sometimes), and “medication-assisted therapy” (53.5% usually or sometimes). Use of preferred terms regarding substance use was common, that is, “drug use” (93%), “substance use disorder” (92.6%), and “addiction” (85.9%), although almost 70% reported use of the nonpreferred term “substance abuse,” and over 30% reported use of “habit” or “clean” at least sometimes. Use of both the preferred term “medications for opioid use disorder” and nonpreferred “medication-assisted therapy” was reported by over half of respondents (60.5% and 53.5%, respectively). Among the 5 nonpreferred terms for a person who uses substances, the most commonly endorsed were “user” (20.2%) and “addict” (15.6%).

FIGURE 1.

Language use related to substance use disorder as categorized by the American. Academy of Pediatrics (2022). A, Represents the use of recommended terms. B, Represents the use of terms that are not recommended. Stacked bar graphs depict the percentage of respondents who report usual use (black), occasional use (dark gray), and no use (light gray).

Provider Attitudes Toward Mothers With OUD

Full verbiage for the 15 attitudinal statements and the distribution of participant responses to them are shown in Figure 2. Overall, nearly all respondents agreed that mothers in treatment for OUD can successfully recover (98.2%), can be a good mother (98.5%) and can successfully raise children (96.7%), and that a mother with OUD has an illness (93.4%). There was greater variability in agreement with statements that a mother with OUD is responsible for her substance use (54.7%), has put her child in danger (56.8%), will follow medical advice (52.0%), will follow through with health care appointments (47.6%), or will understand medical information (43.8%). Most participants disagreed with the statements that a mother with OUD has acted selfishly in becoming pregnant (88.2%) or should not be allowed to retain child custody (80.2%).

FIGURE 2.

Provider attitudes toward mothers with opioid use disorder (OUD). Stacked bar graphs depict the percentage of respondents who disagree with the statement (black), who are neutral toward the statement (dark gray), and who agree (light gray). Brackets with labeling on the right indicate how survey items were grouped based on factor analysis.

Factor analysis was applied to these attitudinal survey items, resulting in a correlation matrix of 5 factors (see Supplemental Table, Supplemental Digital Content 1, http://links.lww.com/JAM/A608 for correlation values). Figure 2 indicates the 4 correlated survey items that assess clinician trust/confidence in mothers in treatment for OUD, therefore comprising a “trust” component score. Also shown are 4 survey items that were strongly correlated in assessing punitive attitudes toward prenatal OUD, comprising a “blame” component score. Finally, 4 survey items were strongly correlated that assess clinician support for mothers in treatment for OUD, comprising a “support” component score. Cumulatively, these 3 component scores accounted for a majority of the variance in data across all attitudinal survey items and were selected as inputs for the next step of analysis (k-means clustering).

Next, cluster analysis evaluated participant responses across each of these main 3 component scores of trust, blame, and support. As shown in Figure 3, participants clustered into 3 distinct groups based on their responses to trust items, 4 groups based on their responses to blame items, and 3 groups based on their responses to support items. Of note, participants in the “high support” cluster were less likely to agree with the statement that “A mother who receives treatment can successfully recover from opioid use disorder,” while almost universally agreeing with the other 3 support variables. In multivariable logistic regression (Table 2), a high level of trust was associated with confidence in training/education to provide care for children with IOE. A low level of blame was associated with personal experience of being impacted by SUD, as well as education/training on the impact of stigma surrounding OUD.

FIGURE 3.

Scatter plot representing clustering of attitudinal responses. Scatter plot of each participant’s mean Likert response responses to “trust,” “blame,” and “support“ statements. A, Depicts 3 “trust“ clusters. B, Depicts 3 “blame” clusters. C, Depicts 3 “support“ clusters. Cluster designation for each participant is represented by symbols.

TABLE 2.

Adjusted Associations Between Participant Characteristics and Attitudinal Clusters, Multivariable Logistic Regression^*

	High trust, n = 99 AOR (95% CI)	Low blame, n = 72 AOR (95% CI)	High support, n = 145 AOR (95% CI)
Clinical role
Attending/nurse practitioner	Reference	Reference	Reference
Resident	1.57 (0.55, 4.48)	0.35 (0.12, 1.04)	1.01 (0.38, 2.65)
Female gender	0.75 (0.41, 1.35)	1.23 (0.63, 2.43)	1.19 (0.66, 2.13)
Age
35 y or under	Reference	Reference	Reference
36–45 y	1.53 (0.48, 4.84)	0.39 (0.11, 1.35)	1.28 (0.44, 3.75)
46–55 y	2.40 (0.69, 8.36)	0.71 (0.19, 2.57)	0.80 (0.25, 2.59)
56 y or older	0.60 (0.12, 3.08)	0.28 (0.05, 1.52)	1.28 (0.32, 5.07)
Personal impact of substance use disorder	1.49 (0.87, 2.58)	2.24 (1.18, 4.26)	1.55 (0.92, 2.60)
Mostly Medicaid payor mix	0.92 (0.52, 1.63)	0.72 (0.39, 1.39)	1.34 (0.78, 2.30)
≥ 6 children with intrauterine opioid exposure seen past 12 mo	0.85 (0.46, 1.57)	0.73 (0.37, 1.45)	0.91 (0.51, 1.64)
Confidence in the adequacy of training	1.84 (1.03, 3.31)	0.97 (0.50, 1.88)	0.98 (0.55, 1.75)
Stigma training	1.50 (0.84, 2.69)	3.43 (1.69, 6.98)	1.64 (0.93, 2.87)
MOUD training	1.16 (0.54, 2.50)	0.79 (0.33, 1.89)	0.70 (0.34, 1.43)
SBIRT training	0.76 (0.44, 1.31)	1.68 (0.92, 3.07)	1.33 (0.78, 2.26)

^*Coefficients represent the adjusted odds ratio (AOR) of assignment to a specified cluster, adjusted for all shown covariates.

Bolded values indicate P value < 0.05. MOUD indicates medications for opioid use disorder; SBIRT, screening, brief intervention, and referral to treatment.

DISCUSSION

This survey-based study conducted across a diverse set of 28 pediatric primary care teaching clinics provides a unique lens into stigma and attitudes toward mothers with OUD among pediatric providers. Overall, respondents reported frequent use of non-stigmatizing SUD-related language and high levels of support for mothers with OUD; however, responses reflected greater variation in attitudes of blame toward and trust in these mothers. Our cluster analysis suggests that exposure to and education about populations with OUD are associated with more positive attitudes toward mothers with OUD among pediatric primary care providers.

In our study, pediatricians reported using preferred substance use–related terminology more often than nonpreferred terminology and rarely using the most obviously stigmatizing words listed (eg, “junkie” and “dirty”). While the importance of stigmatizing language is widely recognized in the literature and by diverse professional societies, including the AAP and other organizations,^20,30–32 few previous studies have characterized language use. Prior studies have demonstrated frequent use of nonpreferred terminology in written text, including news articles, online forums, and patient medical records.^33,34 In particular, the term “substance abuse,” which is nonpreferred relative to “substance use disorder” as it suggests that use is intentional rather than part of a medical condition, has been identified as among the most commonly used stigmatizing terms in clinical notes.³⁴ Similarly, pediatricians in our study report moderate to high frequency of using the nonpreferred term “substance abuse,” as well as other nonpreferred terms, such as “medication-assisted therapy” and “relapse.” These terms may be less obviously pejorative than other nonpreferred terms listed in our survey. Their higher reported use may reflect a lack of exposure to more recent terminology, such as “recurrence of use,” which is felt to be less judgmental than “relapse,” and “medications for OUD,” which appropriately acknowledges medications as the core component of rather than merely “assisting” with OUD treatment.^19,20 Moreover, as the shifting landscape of preferred language is perpetual, this reflects the need for continued education to promote medically accurate terminology and cultural contextualization of care approaches and research results.

Our analysis demonstrated the cohesion of attitudinal survey items into 3 main categories—support, blame, and trust—which parallel the classification of similar survey items in prior research.^16,29 Our finding that the statement “A mother who receives treatment can successfully recover from OUD” is negatively correlated with other supportive statements was unexpected. While the reasons for this negative correlation are not entirely clear, it may reflect a primary conceptualization of substance use disorder as a treatable, chronic disease rather than a curable condition or habit.

Overall, respondents in our study reported high levels of support for mothers with OUD with more mixed attitudes regarding blame and trust, mirroring patterns seen in samples of pediatric trainees and the general population previously surveyed by Schiff et al.^16,29 As in these prior studies, the preponderance of our respondents conveyed belief in the potential for recovery and successful parenting among mothers with OUD. Endorsement of blame-related statements was more variable, though substantially less common among pediatricians in our study than among the general population.²⁹ Few respondents to our survey supported custody loss or agreed that mothers with OUD had acted selfishly, but about half endorsed belief in maternal responsibility for SUD and felt the child had been endangered. Comparatively, within the general population, 33% felt that using opioids during pregnancy was cause for custody loss, and 80%–90% endorsed beliefs in maternal responsibility for SUD and child endangerment. Responding pediatricians also demonstrated some mistrust in the ability of mothers with OUD to appropriately and adequately engage in pediatric care for their child, similar to prior findings among pediatric trainees at a single institution.¹⁶ This pattern of conflictual attitudes may reflect the potential tensions experienced by pediatricians–balancing support of caregivers with their responsibility to protect and promote the health of their pediatric patients.

Given the impact of stigma on patient and family engagement in pediatric care, identifying ways to reduce such stigma among pediatric providers is imperative. The association seen in our study between exposure to and training on SUD and more positive pediatrician attitudes (ie, greater trust, lower blame) highlights education as a potential means to shift attitudes. In fact, prior studies evaluating a wide range of interventions, including SUD-focused clinical rotations, training modules incorporating SUD-related personal narratives, and large-scale campaigns focused on non-stigmatizing language, have demonstrated effectiveness in reducing SUD-related stigma among medical professionals.^35,36 Our results provide further support to the promise of evidence-based education and stigma reduction campaigns and the need for research on their impact on the quality and experience of care for patients and families affected by SUD. To ensure broad reach and maximize impact, educational opportunities should be incorporated into medical school and residency curriculums as well as into continuing medical education offered by pediatric professional societies and licensing bodies.

Our findings should be interpreted in light of both study strengths and limitations. To our knowledge, this is the first large-scale US survey assessing pediatrician attitudes and language use regarding mothers with OUD. However, our findings may lack generalizability outside of academic settings, as all respondents were affiliated with pediatric residency teaching clinics. In addition, response biases inherent to self-reported survey data—particularly social desirability and acquiescence bias—may have increased endorsement of responses portraying positive attitudes toward mothers with OUD. To reduce the risk of response bias, we took steps to ensure participant anonymity and optimize survey formatting (ie, randomizing the order of items, using non-leading questions, and providing “neutral” response options). The survey response rate, with the potential introduction of nonresponse bias, is another limitation of this study. It is possible that clinicians with greater awareness of and interest in maternal OUD were more likely to complete the survey and less likely to use nonpreferred terms or hold negative attitudes toward these mothers. However, prior research indicates that nonresponse bias tends to be low among pediatricians regardless of response rate.³⁷ In line with these findings, our analysis of nonresponses based on measurable characteristics suggests that our analytic sample is reasonably representative of the overall sample of contacted clinicians. Overall, survey response rates have declined in recent years, both among the general public in the setting of the COVID-19 pandemic and more specifically among pediatricians.^37–39 Among health profession trainees, survey response rates tend to vary widely, with lower responses for multi-institution compared with single-institution surveys.⁴⁰ Importantly, our overall response rate of 32.8% to this survey mirrors the response rate for the 2021 AAP Periodic Survey (35.8%) with similar demographic characteristic distribution.³⁹

CONCLUSIONS

Pediatricians demonstrate high levels of support for mothers with OUD but also often endorse blame related to substance use as well as mistrust that mothers with OUD can appropriately engage in pediatric care. While pediatricians tend to use preferred substance use–related terminology and avoid many obviously stigmatizing terms, they sometimes use language that implies blame for substance use and may be medically inaccurate. Training on stigma and additional exposure to strategies to address parental SUD may reduce negative attitudes among pediatricians and potentially improve outcomes for families affected by SUD.

Supplementary Material

Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.journaladdictionmedicine.com.