Obsessive-Compulsive Disorder Misdiagnosis among Mental Healthcare Providers in Latin America

Abstract

Mental healthcare professionals often have limited awareness of different obsessive-compulsive disorder (OCD) symptom presentations, which may contribute to years between OCD symptom onset and treatment initiation. While research has identified high rates of OCD misdiagnosis among clinicians from the United States and Canada, research on OCD symptom awareness among healthcare providers in Latin American (LATAM) regions is limited. In this study, LATAM mental healthcare providers (N = 83) provided diagnostic impressions based on five OCD vignettes: three with symptoms centered on taboo thoughts (sexual, harming others, and religion/scrupulosity) and two about contamination or symmetry obsessions. Rates of incorrect (non-OCD) diagnoses were significantly higher for the taboo thoughts vignettes (sexual, 52.7%; harm/aggression, 42.0%; and religious, 34.7%) vs. contamination obsessions (11.0%) and symmetry obsessions (6.9%). The OCD vignette depicting sexual obsessions was often attributed to a paraphilic disorder (36.5%). Bachelor’s level clinicians had significantly lower odds of accurately identifying all three vignettes related to taboo thoughts compared to respondents with a graduate degree. Accurate identification of the three taboo vignettes was also associated with first-line psychological treatment recommendations (i.e., cognitive-behavioral therapy) even when controlling for respondents’ theoretical orientation. Exposure was rarely mentioned when clinicians were prompted to provide treatment recommendations for each vignette (8–9% of the time for symmetry and contamination vignettes, 5–7% for taboo though vignettes). Like clinicians in the United States and Canada, mental health professionals in LATAM may misidentify OCD symptom presentations, particularly sexual obsessions, highlighting a need for education and training.

1. Introduction

Obsessive-compulsive disorder (OCD) is a heterogeneous condition characterized by obsessions (i.e., recurrent and unwanted thoughts or images) and compulsions or avoidance (i.e., actions to reduce or neutralize obsessions; American Psychiatric Association, 2013). OCD affects approximately 1 to 2% of individuals globally (Zohar, 1999), typically follows a chronic course (Sharma & Math, 2019), and confers significant impairment (Markarian et al., 2010). OCD symptoms are often grouped into four factor analytically derived dimensions: symmetry obsessions and ordering/arranging/counting compulsions, contamination obsessions and hygiene compulsions, obsessions about doubt/harm and checking compulsions, and taboo-related obsessions and compulsions (themes of violence, sex, religion), with many studies finding that the doubt/harm and taboo thoughts categories are more appropriately characterized by a single factor (e.g., Stewart et al., 2008; Bloch et al., 2008). Cognitive-behavioral therapy (CBT) with exposure and response presentation (ERP) and serotonin-reuptake inhibitors (SRIs) are evidence-based treatments that provide clinically significant symptom relief for the majority of people with OCD (Bergez et al., 2020; Öst et al., 2016). Unfortunately, individuals with OCD often experience a significant delay between onset and initiating evidence-based treatment, with a mean delay between onset and any treatment ranging between 10 and 17 years in the United States (see García-Soriano et al., 2014 for a review).

Part of the delay between symptom onset and receipt of evidence-based intervention may be attributed to a lack of healthcare provider awareness of the varying symptom presentations of OCD, particularly taboo-related symptoms. Glazier et al. (2015a) examined 208 U.S. primary care physicians’ ability to recognize various OCD presentations through use of vignettes entailing one of the following common OCD subtypes: aggression, contamination, doubt/fear of saying things, homosexuality, pedophilia, religious, somatic, and symmetry-related concerns, and provided their diagnostic impressions and recommendations. OCD symptoms were misdiagnosed about half (50.5%) of the time overall. Misidentification rates by vignette type were sexual obsessions (70.8 – 84.6%), aggression (80.0%), somatic concerns (40.0%), religion (37.5%), contamination (32.3%), and symmetry (3.70%). Physicians who misidentified OCD vignettes were less likely to recommend empirically supported treatment (cognitive-behavioral therapy [CBT] = 46.7%, selective serotonin reuptake inhibitor [SSRI] = 8.6%; Glazier et al., 2015a), and were more likely to recommend antipsychotic medications (12.4%). Among 360 practicing mental healthcare providers in the U.S., Glazier et al. (2013) found an overall misidentification rate of 38.9% across OCD case vignettes. Providers had substantially more difficulty recognizing uncommon or taboo-themed OCD symptoms (up to ~75% misidentification rate), compared to more common case vignettes such as contamination obsessions (15.8% misidentification rate; Glazier et al., 2013).

Training materials and popular media in the U.S. tends to focus on contamination and symmetry OCD symptoms, giving providers and trainees fewer opportunities for training and discussion about taboo-themed OCD obsessions (Glazier & McGinn, 2015). Indeed, difficulty recognizing taboo obsessions has been documented among graduate student clinicians in U.S. clinical, counseling, and school psychology programs. Among 82 advanced doctoral students, taboo obsessions were misdiagnosed up to 36% of the time, compared to less than 5% for contamination and symmetry OCD obsessions (Glazier & McGinn, 2015).

There are important clinical implications for misdiagnosis. Beyond sustained disability and distress, misdiagnosis of OCD leads to inappropriate treatment decisions (Maršanić, Aukst-Margetić, Grgić, & Kušmić, 2011). Furthermore, individuals experiencing taboo obsessions may be likely to fear stigmatization and being misunderstood by their providers. Taboo obsessions, for example involving intrusive thoughts related to violence, rape, or other socially unacceptable acts, are more likely to produce feelings of shame (Glazier et al., 2015b), which may, in turn, make an individual less likely to voice their distress to family members, seek out information, and seek treatment. If providers do not recognize taboo obsessions if they are disclosed by their patients, this could contribute to even more shame and self-stigma. Unfortunately, little research has examined what factors are associated with better OCD identification and subsequent therapy recommendations. Further research is needed to understand patterns of OCD misdiagnosis, particularly among Latin American (LATAM) populations.

To date, studies have not addressed whether misdiagnosis patterns found by Glazier et al. (2013; 2015a) extend to other populations, particularly LATAM clinicians, and in regions where individuals may have limited access to specialist-care. Accordingly, the present study sought to better understand mental health providers’ diagnostic impressions and therapy recommendations for individuals with OCD in LATAM regions. Specifically, the extent to which providers in LATAM countries distinguished between types of obsessive-compulsive symptoms (i.e., ‘common’ and ‘taboo’ symptom clusters) using a vignette-based approach was examined. First, we aimed to identify the rate of accurate OCD identification among LATAM mental health providers. Similar to misdiagnosis trends in the U.S., we hypothesized that OCD would be more likely to be misidentified when providers are presented with vignettes of individuals with taboo versus contamination and symmetry-themed OCD symptoms. Second, we sought to examine treatment recommendations for each vignette, specifically evaluating how often evidence-based psychological therapy recommendations were made. Finally, we examined cultural and demographic variables impacting misdiagnosis rates for the OCD case vignettes. We hypothesized that a cognitive-behavioral theoretical orientation, self-reported confidence working with patients with OCD, more years of clinical experience, and more advanced education/clinical training background would be associated with improved diagnostic accuracy across each taboo vignette. We also investigated if accurate diagnosis would be associated with evidence-based psychological treatment recommendations (i.e., cognitive-behavioral therapy).

2. Methods

2.1. Participants

Participants included various types of Spanish-speaking mental health clinicians (therapists, psychologists, psychiatrists, etc.) who were actively treating individuals in Mexico, Central America, and South America. In total, 112 individuals between the ages of 22 – 70 years old initiated the study survey, with 74.1% (N = 83) participants successfully completing at least one contamination or symmetry case vignette, and at least one taboo OCD case vignette. After assessing eligibility for those who completed the survey, data from respondents who completed at least one contamination or symmetry vignette, and at least one taboo OCD vignette, were used for analysis (N = 83). Countries represented included Mexico, Panama, Costa Rica, Colombia, Guatemala, Ecuador, Peru, Uruguay, Argentina, Paraguay, Chile, Brazil, Nicaragua, El Salvador, Honduras, Panama, Venezuela, Guayana, and Bolivia, with the largest portion of participants indicating that they were from Mexico (20.5%). 67.5% of the sample was female. 20.5% of providers had obtained a bachelor’s degree, 50.5% master’s degree, and 28.8% doctorate degree. In terms of professional identity, the majority of the present sample reported employment as psychologists (bachelors through doctoral degrees; 80.4%), followed by psychiatrists (12.0%), and family therapists/counselors (8.4%). In the present sample, the mean number of years working in a mental health setting was 14.34 years (SD = 10.02, range = 1 – 40 years in practice). Please see Table 1 for a summary of demographic characteristics of respondents.

Table 1.

Demographics and Clinical Experience of Survey Respondents (N = 83)

Characteristic
Age, mean (SD) years	40.29 (10.57)
Female gender, %	67.5
Hispanic or Latino(a), %	94.0
Race, %
White	48.2
Clinical Experience
Years in clinical practice, mean (SD)	14.34 (10.02)
Currently practicing, %	97.0
Highest Level of Education, %
Bachelor’s Degree	20.50
Master’s Degree	50.53
Doctorate	28.76
Mental Health Profession, %
Psychologist	80.7
Neuropsychologist	1.2
Psychiatrist	12.0
Family Therapist or Counselor	7.2
Country of Residence, %
Argentina	9.6
Bolivia	4.8
Brasil	1.2
Chile	2.4
Colombia	4.8
Costa Rica	2.4
Ecuador	1.2
El Salvador	9.6
Guatemala	7.2
Honduras	4.8
Mexico	20.5
Nicaragua	2.4
Panama	2.4
Paraguay	4.8
Peru	1.2
Venezuela	4.8
Uruguay	6.0
Missing	9.6
Professional Setting, % a
University	42.2
Private Practice	80.7
Hospital	15.7
Mental Health Facility	13.3
School Setting	6.0
Other	16.9
Populations Treated, %
Children/Adolescents	22.9
Adults	56.1
Children and Adults	19.5
Theoretical Orientation, %
Cognitive-Behavioral	39.8
Cognitive	1.2
Behavioral	2.4
Psychoanalytic/Psychodynamic	19.3
Interpersonal	3.6
Humanistic/existential	8.4
Acceptance and Commitment Therapy (ACT)	3.6
Other	21.7

^aDenotes more than one answer was permitted.

2.2. Procedure

The study was approved by the Baylor College of Medicine Institutional Review Board (IRB). Given the utility of email communication strategies for recruiting specific populations of interest (King, O’Rourke, & DeLongis, 2014), participants were primarily recruited by using publicly available emails from university, hospital and clinic websites, and LinkedIn, amongst others. A total of 965 emails were sent out to providers or mental health clinics, out of which 895 were successfully received. The emails that were sent out to the mental health providers included a brief explanation of the study, a flyer with inclusion criteria, and the link to access the survey. Approximately one month after sending a first e-mail, a second email was sent as reminder to complete and/or help distribute the survey. Additionally, 35 providers/clinics were contacted through social media, for a total of 1,000 clinics/providers contacted directly. Furthermore, for indirect recruitment, a one-time Facebook event was created, as well as a post on Twitter and Instagram. Participants were encouraged to distribute the survey to fellow colleagues as a secondary method of recruitment. Recruitment and data collection took place over a 3.5-month period, from mid-March 2020 through June 2020. Participation was voluntary, responses were collected anonymously, and no compensation was provided.

2.3. Survey Instrument

A ~30-minute self-report survey was administered to participants on Qualtrics. First, participants answered demographic questions. Second, participants were presented with nine vignettes depicting psychiatric clinical cases presented in random order (see below description). Participants were then asked to select a diagnosis from a list of choices and a free-response recommended route(s) of treatment for each case. Participants were also asked about number of years in clinical practice, theoretical orientation, and self-reported confidence with diagnosing OCD. These questions were presented after vignettes to avoid biasing OCD diagnoses in the earlier part of the survey.

Modeling from and extending the work of Glazier et al. (2015), the nine vignettes assessed mental health professional’s accuracy in identifying specific symptom presentations of OCD (five vignettes) versus other non-OCD (four vignettes) psychiatric diagnosis. Three of the OCD vignettes focused on taboo symptom presentations: sexual obsessions (involving intrusive thoughts about family and/or children), religious obsessions (scrupulosity), and obsessions related to causing others harm¹. Two of the OCD vignettes focused on well-known presentations of OCD: contamination and symmetry. The non-OCD cases presented a patient with one of the following psychiatric disorders: anxiety disorder, body dysmorphic disorder (BDD), schizophrenia, and major depressive disorder (MDD).

To reduce content bias, the age of the patient described in each of the nine vignettes remained constant. The length of vignettes was one paragraph (word count range: 89–126). The content of the nine vignettes were developed and validated by English-speaking OCD specialists. Vignettes were translated to Spanish and then back translated to English to ensure translational equivalence. Translations were conducted by two Spanish-speaking masters-level researchers with OCD experience and were then reviewed and approved by an independent native Spanish-speaking masters-level researcher.

2.4. Data Analysis

Data were analyzed using SPSS version 25. The content of written treatment recommendations provided by respondents was conducted by trained research coordinators under the supervision of experienced OCD clinicians/researchers. Freely written responses for each treatment recommendation vignette were interpreted from Spanish to English, then scored for whether they reflected first-line therapy recommendations (i.e., CBT; Koran et al., 2007). Although ERP is a specific form of CBT that has the strongest evidence in the treatment of OCD, due to the open-ended nature of the question and the common shorthand of CBT to include ERP, as well as following prior work in this area (Glazier et al., 2015a), we chose to include any response that included “CBT” as accurate, though we also present data on whether exposure was described in a free-response item probing treatment recommendations. We also chose to focus on therapy but not pharmacotherapy recommendations considering that the majority of the sample identified as psychotherapists (i.e., not prescribing physicians).

Next, rates of accurate OCD identification and treatment recommendations were presented. To investigate whether OCD misdiagnosis was greater for taboo vignettes than contamination/symmetry, McNemar proportion difference tests were used to compare rates of misdiagnosis for each taboo vignette with contamination and symmetry vignettes. Because there were very few prescribing providers in this sample (n = 12%), we chose to focus on psychological treatment recommendations.

The final set of analyses evaluated cultural and demographic variables associated with misdiagnosis (OCD vs. other) within the three taboo thought vignettes. Specifically, misdiagnosis and CBT recommendations were compared with chi-square proportion difference tests among therapists with cognitive-behavioral (including those who identified as cognitive-behavioral, cognitive, and behavioral) and other orientations, as well as among individuals with a graduate and with a bachelor’s degree. Independent samples t-tests were used to compare years of experience and self-reported confidence with OCD among individuals who did and did not accurately identify the vignette correctly. All variables had skewness and kurtosis values less than two, suggesting appropriate distributions for parametric analyses.

3. Results

3.1. OCD misidentification by vignette type

The symmetry obsessions vignette resulted in the lowest misidentification rate (7%), followed by contamination OCD (11%). Table 2 shows the OCD misidentification rates by vignette type. Paraphilic/sexual obsessions were misidentified 53% of the time, the aggression/harm vignette was misidentified 42% of the time, and religious obsessions were misidentified 35% of the time.

Table 2.

Clinical Case Vignette Misidentification Rates

	N	Misidentification Rate n (%)	Most Common Misdiagnosis (n, %)
I. OCD Vignettes
Obsessions regarding...
Contamination	73	8 (11%)	Hypochondriasis (n=4, 6%)
Symmetry/Exactness	72	5 (7%)	Personality Disorder (n=2, 3%)
Aggression/Harm	69	29 (42%)	Anxiety Disorder (n=10, 15%)
Scrupulosity/Religion	75	26 (35%)	Personality Disorder (n=9, 12 %)
Sexual Obsessions	74	39 (53%)	Paraphilic Disorder (n=27, 37%)
II. Non-OCD Vignettes
Depression	70	6 (9%)	Bipolar Disorder (n=3, 4%)
Social Anxiety Disorder	72	19 (26%)	Personality Disorder (n=19, 8%)
Psychosis	70	20 (29%)	Personality Disorder (n=11, 16%)
Body Dysmorphic Disorder	71	5 (7%)	N/A

Note. Only 5 participants misdiagnosed the BDD vignette. Misdiagnoses were: anxiety (N = 1), personality disorder (N = 1), hypochondriasis (N = 1), OCD (N = 1) and “low self-esteem” (N = 1).

Using McNemar’s test for repeated proportion comparisons, the proportion of correct OCD identification from the contamination vignette was found to be significantly greater than the proportion of correct identification for the paraphilic/sexual obsessions vignette, the aggression/harm vignette, and the religious vignette, all p < .001. This was also the case for the symmetry vignette compared to others, all p < .001.

Among participants who endorsed a non-OCD response, the most common incorrect primary diagnoses (listed with the correct OCD type in parentheses) were as follows: anxiety disorder (35%; aggressive obsessions), hypochondriasis (50%; contamination obsessions), personality disorder (27%; religious obsessions), paraphilic disorder (69%; sexual obsessions), and personality disorder (40%; symmetry obsessions). Aggression/harm obsessions were also commonly misdiagnosed as posttraumatic stress disorder (PTSD; 17%).

As a point of comparison, non-OCD vignette misidentification fell into two distinct categories: 2 non-OCD vignettes had low misidentification rates (less than 10%), while the other two non-OCD vignettes had moderately elevated misidentification rates (over 25%). Social anxiety disorder and psychosis were each most commonly misdiagnosed as personality disorders (32% and 55% misidentified as personality disorder, respectively). BDD had the lowest misidentification rate (7%). MDD had a misidentification rate of 9%.

3.2. Treatment recommendations

All providers recommended psychological treatment for each OCD vignette, except for one who did not recommend treatment for the sexual/religious vignette. Although CBT was recommended more than any single other form of therapy for each vignette when evaluating free responses, ERP was specifically mentioned only 5–9% of the time across vignettes (5–7% in taboo sections, 8–9% in contamination/symmetry). Please see Table 4 for a summary of treatment recommendations.

Table 4.

Comparisons of OCD identification for different taboo-themed OCD vignettes

			Test statistic	Effect size a OR [95% CI] or d
Sexual vignette
	Incorrectly identified (n=39)	Correctly identified (n=35)
Theoretical orientation N (%)			χ2 (1) = 1.91	1.92 [0.91, 4.06)
Cognitive-behavioral (n=36)	16 (44%)	20 (56%)
Other (n=38)	23 (61%)	15 (40%)
Education N (%)			χ2 (1) = 12.51**
Bachelor’s degree (n=16)	13 (81%)	3 (19%)		0.19 [0.048, 0.72]
Master’s degree (n=37)	21 (57%)	16 (43%)		1.39 [0.55, 3.46]
Doctoral degree (n=21)	5 (24%)	16 (76%)		5.73 [1.81, 18.09]
Recommended CBT			χ2 (1)= 13.26***	6.61 [2.64, 16.59]
Yes (n=41)	14 (34%)	27 (66%)
No (n=31)	24 (77%)	7 (23%)
Confidence with OCD assessment M (SD)	7.4 (1.9)	8.3 (1.5)	t (62) = 2.14*	.54
Years of experience M (SD)	12.2 (8.4)	16.1 (10.9)	t (71) = 1.72	.40
Aggression/Harm vignette
	Incorrectly identified (n=29)	Correctly identified (n=40)
Theoretical orientation N (%)			χ2 (1) = 0.18	1.23 [0.57, 2.65]
Cognitive-behavioral (n=33)	13 (39%)	20 (61%)
Other (n=36)	16 (44%)	20 (56%)
Education			χ2 (1) = 6.60*
Bachelor’s degree (n=15)	10 (67%)	5 (33%)		0.27 [0.081, 0.91]
Master’s degree (n=33)	14 (42%)	19 (58%)		1.03 [0.40, 2.69]
Doctoral degree (n=21)	5 (24%)	16 (76%)		3.20 [1.01, 10.13]
Recommended CBT			χ2 (1) = 3.83+	2.73 [1.20, 6.26]
Yes (n=34)	11 (32%)	23 (68%)
No (n=30)	17 (57%)	13 (43%)
Confidence with OCD assessment M (SD)	7.7 (1.8)	8.0 (1.7)	t (60) = 0.40	d = .22
Years of experience M (SD)	11.7 (7.6)	16.3 (10.7)	t (66) = 1.94	d = .48
Religion vignette
	Incorrectly identified (n=26)	Correctly identified (n=49)
Theoretical orientation N (%)			χ2 (1) = 0.15	1.21 [0.57, 2.55]
Cognitive-behavioral (n=34)	11 (32%)	23 (68%)
Other (n=41)	15 (37%)	26 (63%)
Education			χ2 (1) = 5.33
Bachelor’s degree (n=15)	9 (60%)	6 (40%)		0.26 [0.081, 0.85]
Master’s degree (n=38)	11 (29%)	27 (71%)		0.60 [0.23, 1.56]
Doctoral degree (n=22)	6 (27%)	16 (73%)		1.62 [0.54, 4.81]
Recommended CBT			χ2 (1) = 13.97***	7.29 [3.12, 16.98]
Yes (n=43)	7 (16%)	36 (84%)
No (n=29)	17 (59%)	12 (42%)
Confidence with OCD assessment M (SD)	7.1 (2.0)	8.2 (1.5)	t (61) = 2.31*	.62
Years of experience M (SD)	13.2 (10.5)	15.3 (10.1)	t (72) = 0.86	.21

⁺p = .05

^*p < .05

^**p < .01

^***p < .001

^aORs in the education comparison were calculated by comparing the odds of correct identification in one educational group vs. the two others pools (e.g., doctoral proportion of correct identification vs. combined bachelor’s and master’s correct identification)

3.3. Factors associated with OCD identification

Please see Table 3 for a summary of factors associated with correct taboo OCD vignette identification. Bachelor’s level therapists were significantly less likely to accurately diagnose all three taboo vignettes than providers with graduate training, whereas those with a doctoral degree were more likely to accurately diagnose the sexual and aggression/harm vignettes compared to those with a master’s or bachelor’s degree. Those who recommended CBT were significantly more likely to have correctly identified each taboo vignette as reflecting OCD. Cognitive-behavioral therapists and therapists who held other theoretical orientations were not found to be different in their diagnostic accuracy. Years of experience in clinical practice was not associated with accurate diagnoses. Individuals who correctly identified the sexual and religious vignettes self-reported more confidence with OCD diagnostic assessment.

Table 3.

Clinical Case Vignette Treatment Recommendations

		Recommendation n (%)
	N	Psychological Treatment	Exposure and response prevention	CBT without mention of ERP	Non-Evidence-Based OCD Treatment
OCD Vignettes
Contamination	72	72 (100%)	9 (13%)	41 (57%)	21 (29%)
Symmetry/ Exactness	72	72 (100%)	8 (11%)	41 (57%)	22 (31%)
Aggression/ Harm	68	68 (100%)	5 (7%)	29 (43%)	30 (44%)
Scrupulosity/Religion	75	74 (98.7%)	5 (7%)	38 (56%)	30 (36%)
Sexual Obsessions	74	74 (100%)	7 (9%)	34 (46%)	31 (42%)

Note. Percentage of psychological treatment recommendations add to less than 100% in some columns due to missing responses for specific psychological treatment recommendations after endorsing that they would recommend psychological therapy.

In light of the significant associations between correct CBT treatment recommendations and accurate OCD identification, multivariate logistic regressions were run to evaluate if correct identification was associated with CBT recommendation when controlling for whether the therapist held a cognitive-behavioral/cognitive/behavioral orientation. Across vignettes, correct OCD identification was significantly associated with CBT recommendations when controlling for CBT orientation. See Table 5 for a summary.

Table 5.

Logistic regression predicting CBT recommendation

	Sexual (n=72)			Harm/Aggression (n=64)			Scrupulosity/Religious (n=72)
	OR	R2	χ2	OR	R2	χ2	OR	R2	χ2
Model summary statistics		.27	22.89***		.26	19.22***		.29	24.47***
Cognitive-behavioral orientation	5.25**			9.03***			6.57**
Accurate identification of vignette as OCD	6.28**			3.43*			9.64***

^*p<.05

^**p<.01

^***p<.001

Note: CBT=cognitive-behavioral therapy R² reflects the Cox & Snell R Square estimate

4. Discussion

This study investigates diagnostic accuracy of OCD among LATAM providers, extending previous findings to a socioeconomically and geographically diverse sample. Consistent with mental health professionals from the U.S. and Canada (Glazier et al., 2013, 2015a), OCD misdiagnosis rates were high among LATAM providers when presented with vignettes reflecting themes of taboo/unacceptable thoughts. Over 50% of clinicians misdiagnosed sexual obsessions, 42% of clinicians misdiagnosed symptoms related to aggression/harm, and 35% misdiagnosed religious obsessions and compulsions. As predicted, the three taboo thoughts vignettes were misidentified at a significantly higher rate than the contamination and symmetry vignettes. These missed diagnoses have meaningful clinical implications; misdiagnosis was significantly associated with CBT recommendations, the first-line treatment for OCD, even when controlling for whether clinicians held a CBT theoretical orientation.

The sexual OCD vignette was often attributed to a paraphilic disorder (36.5%), a diagnosis which could exacerbate sexual obsessions, increase self-stigma, deter future help-seeking, and lead to potentially harmful treatment recommendations. In particular, being diagnosed with a paraphilia and treated by a provider as such could lead to continued misattributions of intrusive thoughts (i.e., that sexual obsessions truly indicate the patient may harm someone else) and increase avoidance behavior or compulsions (e.g., increase reassurance-seeking or avoiding places with potential triggers), in turn leading to a worsening pattern of symptoms. Regarding aggression and religious obsessions, misdiagnosis rates were slightly higher in our sample compared to past vignette studies examining doctoral students in clinical, counseling, and school psychology programs in the U.S. (i.e., Glazier & McGinn, 2015), and lower compared to past work examining misdiagnosis rates among U.S. mental health professionals (Glazier et al., 2013) and primary care physicians (Glazier et al., 2015a). It is worth noting that the harm/aggression vignette was most frequently misidentified as an anxiety disorder, which would likely lead to very similar clinical decisions as an OCD diagnosis, compared to disorders that were far more dissimilar and associated with different treatment recommendations for the other two taboo vignettes (i.e., paraphilia for OCD related to sexual thoughts and personality disorder for symptoms related to religious scrupulosity).

Although OCD phenomenology is similar across cultures with few differences in its severity and comorbidity with other psychiatric disorders (Fontenelle et al., 2004; Medeiros et al., 2017; Chavira et al., 2018), some have suggested that OCD might manifest differently based on cultural factors (e.g., Nicolini et al., 2017). For instance, a study by Fortenelle and colleagues (2004) revealed that aggressive obsessions were more frequently reported in Brazil and Middle Eastern countries, possible due to high prevalence of crimes in these regions. Another study by Okasha and colleagues (1994) postulated that religious obsessions were more common in their Egyptian sample due to the emphasis on religion in the country. Thus, taboo (religious, aggression, sexual) obsessions may be not only a highly prevalent but also culturally sensitive topic in Latin America, further extending the importance of our results.

As hypothesized, level of educational attainment and self-reported confidence in diagnosing OCD were directly associated with accurate OCD identification. Specifically, providers with a bachelor’s degree had lower odds of correct OCD identification for each vignette compared to those with graduate degrees, while doctoral-level clinicians had a higher odds of correct OCD identification for the sexual and harm/aggression vignette compared to those with master’s or bachelor’s degrees. In contrast, a cognitive-behavioral orientation and years of experience were not associated with diagnostic accuracy. OCD-focused education on taboo OCD symptom presentations, common differential diagnoses, and ERP for OCD is warranted, and may be particularly important at the undergraduate level where many clinicians in LATAM receive their clinical training.

CBT was significantly more likely to be recommended across taboo vignettes when therapists accurately identified the vignette as OCD, even when controlling for whether respondents’ held a CBT theoretical orientation. This further underscores the importance of increased education regarding these underrecognized expressions of OCD. Clinicians with knowledge about less recognized OCD presentation also appeared to be knowledgeable about empirically supported interventions for this diagnosis, and thus OCD education should continue to emphasize both evidence-based treatments as well as assessment of taboo thoughts and associated behaviors.

An encouraging finding was that across vignettes, CBT was the most recommended treatment approach. That said, exposure-based approaches were rarely explicitly recommended for any of the vignettes. It is possible that respondents incorporated exposure within their framework of CBT, though the low rate of describing ERP explicitly may reflect the well-documented finding that clinicians often do not use exposure-based approaches for individuals with OCD, even when they state that they are “doing CBT” (Senter et al., 2021).

Our findings should be considered in the context of several limitations. First, the survey length may have discouraged individuals from completing all nine vignettes. Second, we surveyed clinicians from a wide range of countries, and thus it was not possible to examine findings within specific countries, which may identify areas for targeted future work in this area. In most LATAM countries, a bachelor’s degree in psychology allows for providers to practice as a psychotherapist. 80% of our sample identified as psychologists, highlighting a difference in training and licensure guidelines between the U.S. and most LATAM regions. The mental health care system is also very different in Latin America with limited funding and lack of professionals (Caldas de Almeida & Horvitz-Lennon, 2010), which should be further taken into consideration when interpreting our results.

Within these limitations, this study indicates that while accurate diagnosis of contamination and symmetry-related OCD was common, OCD vignettes were misdiagnosed over a third of the time when presenting harm/aggression and religion subtypes, and over half of the time when presenting OCD related to sexual wrongdoing. These data speak to the need to extend OCD education efforts to LATAM countries (as well as those in North America), particularly among bachelor’s level clinicians. While considerable efforts have been made in providing training in North America though the International OCD Foundation (e.g., Behavior Therapy Training Institute [BTTI]; Reese et al., 2019), extending these efforts to LATAM countries may help improve diagnostic accuracy as well as availability of ERP treatment. Personalized training delivery models that are culturally sensitive, and balance differences in help-seeking and illness presentation, should be considered.

Highlights.

• In this study, LATAM mental healthcare providers provided diagnostic impressions based on OCD vignettes.

• Rates of incorrect (non-OCD) diagnoses were significantly higher for the taboo thoughts vignettes vs. contamination obsessions and symmetry obsessions.

• Bachelor’s level clinicians had significantly lower odds of accurately identifying vignettes related to taboo thoughts compared to respondents with a graduate degree.

• Accurate identification of vignettes was associated with first-line psychological treatment recommendations.

• Like clinicians in the US and Canada, mental health professionals in LATAM may misidentify OCD symptom presentations, particularly sexual obsessions, highlighting a need for education and training