Men's Dropout From Mental Health Services: Results From a Survey of Australian Men Across the Life Span.

While increasing numbers of Australian men are accessing mental health services, the sustainability of their therapy engagement varies significantly, with many men being lost to follow-up. The current study investigated dropout rates in a large community-based male sample to highlight the reasons for, and potential predictors of, men dropping out of mental health care services. Data were drawn from an online survey of 1907 Australian men (aged 16-85; M = 44.1 years) reflecting on their broad experiences in mental health therapy. Participants responded to bespoke items assessing their past dropout experience and reasons for dropping out, the odds of which were modeled in relation to demographics and predictors (e.g., therapist engagement strategies, alignment to traditional masculinity and pre-therapy feelings of optimism, shame, and emasculation). The overall dropout rate from therapy was 44.8% (n = 855), of which 26.6% (n = 120) accessed therapy once and did not return. The most common reasons for dropout were lack of connection with the therapist (54.9%) and the sense that therapy lacked progress (20.2%). Younger age, unemployment, self-reported identification with traditional masculinity, the presence of specific therapist engagement strategies, and whether therapy made participants feel emasculated all predicted dropout. Current depressive symptoms and suicidality were also higher amongst dropouts. Therapists should aim to have an honest discussion with all clients about the importance of therapy fit, including the real likelihood of dropout, in order to ensure this does not deter future engagement with professional services.

While men and women have been reported to experience mental health issues at equal rates (Smith et al., 2018), men are a third less likely to access help when they need it (Coates et al., 2019; Harris et al., 2015; Yousaf et al., 2015). The prevailing narrative that men are loathe to seek help has stimulated empirical work exploring men’s attitudes, intentions, and experience of stigma when it comes to engaging with mental health services (Addis & Mahalik, 2003). Comparatively, little attention has been paid to the fact that more Australian men are now accessing care than ever before (Harris et al., 2015). Striving to understand what happens to the growing number of men that access mental health care services is of utmost importance given that up to 60% of the men who die by suicide have sought help in the year prior (John et al., 2020; Stene-Larsen & Reneflot, 2019). Describing men’s experiences in therapy, regarding what does and does not engage them, can inform tailored services to meet the needs of this group (Seidler, Rice, River, et al., 2018).

In accessing mental health care, men are often overcoming attitudinal, cultural, and structural barriers linked to masculine socialization that frame help-seeking as an embodiment of weakness, vulnerability, and dependence (Addis & Mahalik, 2003; Rice et al., 2020). Past work has aimed at increasing men’s likelihood of accessing care, likely because the prevailing findings suggest that men’s outcomes will improve if therapy is effectively engaging (Ogrodniczuk, 2006; Staczan et al., 2017). Enduring challenges remain regarding strategies for engaging men with therapy in sustainable ways to optimize their outcomes (Seidler, Rice, Oliffe, et al., 2018). While men are less likely to access therapy in the first place, when they do, they are also often more likely to attend fewer sessions and drop out from therapy prematurely (i.e., disengage from an agreed-upon service before their issues are resolved) compared with women, limiting the efficacy of therapy (Reneses et al., 2009; Seidler, Rice, Dhillon, et al., 2020; Zimmermann et al., 2017). There is a dearth of information regarding reasons for, and predictors of, mental health therapy dropout in males, and insights to which men are more likely to disengage from therapy prematurely and why, are key to effectively tailoring therapy to men (Seidler, Rice, Ogrodniczuk et al., 2019).

Understanding Dropout

Data from outpatient populations suggest that approximately one in four individuals who seek therapy for mental health problems drop out prematurely (Hoge et al., 2014; Reneses et al., 2009; Wang, 2007). There is nevertheless substantial heterogeneity in rates of reported dropout across studies ranging from 0% to 70% (Cooper & Conklin, 2015; Hans & Hiller, 2013; Swift & Greenberg, 2012; Wierzbicki & Pekarik, 1993). Similarly, within these dropout studies, there are also mixed findings regarding any differences between men and women. This variance is likely due in part to varied definition and operationalization of dropout between studies, directly impacting the rates, risk factors, and within-therapy predictors of dropout (O’Keeffe et al., 2019). Commonly utilized definitions of dropout are derived from therapist judgment, where dropout occurs when the client ceases attending an agreed-upon course of care, despite a recommendation for ongoing therapy (Warnick et al., 2012). These definitions privilege a therapist-centric perspective and can be limited by subjective bias (Wierzbicki & Pekarik, 1993). Placing the perspective of therapists above that of the client in defining dropout obscures clients’ agency in their own determination of therapy duration and outcome. This consideration seems especially pertinent in therapy with men, given shared control and decision-making are emerging as essential strategies for male engagement (Beel et al., 2018; Seidler, Rice, Ogrodniczuk, et al., 2018). Existing research examining dropout often disregards the personalized account and autonomy of the client and their reasons for leaving therapy prematurely. Progress in operationalizing dropout has been made, with recent work defining dropout according to whether the termination of therapy is mutually discussed and agreed upon between client and therapist (Hatchett & Park, 2003; O’Keeffe et al., 2019). Researching dropout therefore necessitates assessment of whether clients discuss discontinuing therapy with their therapist prior to terminating, in order to effectively delineate between clients who discontinue following discussion with their therapist and those who drop out without first discussing it with their therapist, also known as unilateral termination (Westmacott et al., 2010). This level of specificity in assessment of dropout has rarely been achieved in the literature to date and is nonexistent when considering the dropout experiences of men.

Impact of and Risk Factors for Dropout

The impact of dropout from therapy can be the risk of future deferral, mistrust, and outright avoidance in future when the need arises (Chen et al., 2017). This is thought to occur due to the individual’s dissatisfaction, negative attitudes, and experiences that often accompany dropout (Richards & Bedi, 2015; Seidler, Rice, Kealy et al., 2020). The implications of this are far-reaching, manifesting in decreased cost-effectiveness and heightened risk of increased morbidity and mortality for men experiencing dropout due to provision of dissatisfactory care (Barrett et al., 2008; Calear et al., 2014).

While specific risk factors for dropout among men are poorly understood, specific client characteristics are consistently discussed as predictors of dropout, albeit with small effect sizes (Hans & Hiller, 2013). Younger age, unemployment, low income, ethnic minority status, lower educational attainment, more severe symptoms, and doubt regarding the effectiveness of therapy have consistently been linked to greater risk of dropout among mixed-gender samples (Edlund et al., 2002; Egan & Kenny, 2005; Henzen et al., 2016; Linardon et al., 2019; Seidler, Rice, Dhillon, et al., 2020; Wang, 2007). Whilst commonly framed as barriers to entry into mental health services, clients’ lack of motivation, pessimism as to the likely outcome of attending therapy, and experiential shame in attending therapy could serve as barriers to engagement and also predict dropout (Edlund et al., 2002). Given suggestions that psychotherapy can represent a violation of the doctrines of masculinity for men (Westwood & Black, 2012), understanding the extent to which feelings of emasculation on seeking therapy (i.e., feeling like “less of a man”) may be linked with dropout is an essential area for further inquiry. Gaining a more nuanced understanding of predictors of dropout among men will help to equip therapists to appropriately deliver mental health therapy that is sensitized to their needs.

Current Study

The current study objective was to unpack men’s mental health therapy dropout from diverse services (i.e., therapy, psychotherapy, etc.). We first explored rates of, and reasons for, past dropout among a community sample of Australian men, followed by examining predictors of dropping out due to various reasons, according to established demographic predictors in the literature. It was hypothesized that rates of dropout would be higher among younger men, those with lower education levels, and those experiencing unemployment. In addition, the role of masculinity, pre-therapy attitudes and expectations, and experiences in therapy were also explored as predictors of dropout according to various reasons for dropout identified by participating men.

Method

Procedure

Participants were recruited via social media advertisements; inclusion criteria were for men in Australia aged 16 and over with experience in mental health therapy to complete a brief survey about their experiences. The survey was delivered via Qualtrics, and an initial automatic screen was put in place to prevent individuals residing outside of Australia from accessing the survey. The advertisements were presented via Facebook, delivered via social media channels administered by Movember—the world’s leading men’s health charity. Social media advertisements included the following text: “Have you ever had counselling or therapy? If you can spare 15 minutes, we’d love to hear from you so we can improve therapy for men.” Participants who clicked through the initial advertisements were presented with brief information about the survey (i.e., ~15 min to complete, aiming to understand their experiences in therapy), followed by a plain language document explaining the nature of the study, and an online consent form. Participants then responded to a series of questions examining their experiences in therapy, along with certain standardized measures. Participants were given the option to enter the prize draw to win one of fifty $100 gift cards to compensate for their time. Ethics approval was obtained from the University of Melbourne Human Ethics Sub-Committee (study approval number: 1956099).

Measures

Demographics

Individual items assessed participants’ age, sexuality, location (i.e., living in a metropolitan, regional or rural location), relationship status, employment status, and highest level of education obtained.

Dropout

Survey items were developed in order to assess participants’ experiences of dropout from therapy, applying the currently accepted definition of dropout occurring when the client ceases therapy without consulting with their therapist (O’Keeffe et al., 2019). Participants initially responded to “Have you ever stopped attending therapy prematurely? That is, come for one or a few sessions, and didn’t go back to that therapist?” Participants who endorsed this item were then asked, “Did you discuss this with your therapist prior to stopping?” Participants who reported ceasing therapy without discussing this with their therapist were coded as having dropped out.

Participants also reported whether they were currently engaged in therapy at the time of survey completion, and if not, when their most recent experience occurred. Additionally, respondents were asked to report whether their current or most recent experience was also their first experience of therapy. Participants reporting that their current or most recent experience was also their first, were deemed to only have had a single occasion of care.

Reason for Dropout

Participants reporting dropout were also asked, “What was the main reason you decided to stop attending?” Several response options were provided, derived from past literature, along with an open response option for participants who wished to detail their own reason for dropout (Wang, 2007).

Therapist Engagement Strategies

To understand participants’ experiences of their therapists’ efforts to engage them in therapy, participants were asked to rate whether or not they felt various therapist strategies occurred during their engagement and orientation into therapy. The 13 strategies (i.e., therapist microskills) were derived from past summaries of effective engagement tactics for working with male clients and were included with a view to achieving an overall index of the extent to which participants were engaged in therapy (Beel et al., 2018; Seidler, Rice, Ogrodniczuk, et al., 2018). The items, presented in full in the supplementary material, covered various areas including effective orientation to therapy (e.g., The therapist asked about my expectations of therapy), adopting a strengths-based masculinities perspective (e.g., The therapist talked about me seeking help in positive terms), and shared control and decision-making (e.g., The therapist checked whether I felt like therapy was working for me). In this study, the number of strategies that participants reported occurred according to a dichotomous yes/no scale were summed and used as an overall index of the extent to which therapists attempted to engage and orient men appropriately to therapy, with higher scores indicating greater engagement in therapy.

Masculinity

Identification with masculinity was assessed using the Traditional Masculinity Femininity Scale (TMF; Kachel et al., 2016). Participants were asked to rate various attributes of themselves on a seven-point Likert scale from not at all masculine to totally masculine. Example items include “I consider myself as. . .,” “Traditionally, my interests would be considered as. . .,” and “Traditionally, my behaviour would be considered as. . .” Scores are summed to provide an overall score indicating the extent to which participants identify themselves as traditionally masculine, ranging from 6 to 42, with higher scores indicating greater identification with traditional masculinity. Reliability of the scale was strong in the present sample (α = .87).

Pre-Therapy Items

Participants also rated their expectation of the usefulness of therapy, motivation for therapy, and feelings upon presentation to therapy, using items adapted from past literature (Edlund et al., 2002). Expectation of the usefulness of therapy was assessed with a single item: “Before you started therapy, to what extent did you think it would help?” Motivation for therapy was also assessed with a single item: “Before you started therapy, how motivated were you to attend?” Feelings of emasculation and shame upon presentation to therapy were assessed with two items: “Did going to therapy make you feel like less of a man?” and “To what extent did you feel shame in going to therapy?” All items were rated on a five-point Likert scale from not at all to extremely.

Depression

Current symptoms of depression were assessed using the nine-item Patient Health Questionnaire (PHQ-9; Kroenke et al., 2001). The PHQ-9 is a well-established measure of depressive symptomatology, requiring participants to rate their experience of various symptoms in the 2 weeks preceding assessment, on a four-point Likert scale from not at all to nearly every day. Higher scores indicate increased severity of depression symptoms, ranging from 0 to 27. Item 9 of the PHQ-9 was used to categorize participants experiencing suicidality: “thoughts that you would be better off dead or of hurting yourself in some way,” with participants reporting any response other than not at all deemed to be experiencing suicidality. Reliability of the scale was strong in this sample (α = .91).

Data Analysis

Past dropout status was coded as a dichotomous variable, where participants were classified as a dropout if they reported previous discontinuation of therapy without discussion with their therapist. The overall dropout rate was examined for the whole sample, followed by subgroup analyses to uncover the dropout rate among participants reporting on only one occasion of care.

Univariate analyses involved comparisons of means and frequencies on all measures between the dropout and non-dropout groups, using chi square tests of association for categorical variables, and t-tests for continuous variables. Effect sizes were examined using odds ratios for 2 × 2 categorical variables, and Cohen’s d for continuous variables (d = 0.2 was considered a “small” effect, d = 0.5 a “medium” effect, and d = 0.8 a “large” effect size; Cohen, 1988). Age was entered into analyses as a continuous variable, and sexuality was collapsed into a dichotomous variable with two levels: heterosexual and non-heterosexual.

Next, a multinomial logistic regression was used to identify predictors of specific reasons for dropout. Only those measures identified as significant via univariate tests were included as predictors. The reference category in the regression was “no dropout,” which included participants who had either not dropped out or discussed their discontinuation with their therapist. This allowed an understanding of the odds of dropping out according to specific dropout reasons, relative to the levels of independent variables included. For this analysis, participants who did not report a reason for dropping out (n = 12) were excluded. Model fit was evaluated according to likelihood ratio tests for the overall model and for individual predictors. All statistical analyses were performed using SPSS version 26.

Results

Sample Demographics

In total, 2009 men responded to the survey. A minority of participants reported a current episode of care that was also their first experience (102; 5.1%); these participants were excluded from further analyses as they were deemed to not have had an opportunity to experience dropout. This left a remaining sample of 1907 participants who were included in further analyses. Of the sample, 508 (26.6%) reported they were currently in therapy, 1123 (58.9%) reported their most recent experience was up to 5 years ago, and 276 (14.7%) reported their most recent occasion of therapy was over 5 years ago. A minority of participants were reporting on their first and only experience (451; 23.6%), with most reporting they had experienced multiple occasions of therapy (1456; 76.4%).

Among the 1907 participants, the mean age was 44.12 years (SD = 15.21 years; range 16–85 years). The majority of participants were heterosexual (72.2%, n = 1376), living in a metropolitan area (60.2%, n = 1148), and employed (65.9%, n = 1257). The TMF scale mean for the sample was 28.7 (SD = 6.21; range 6–42). The mean PHQ-9 score was 10.0 (SD = 6.88, range 0–27), with current suicidal ideation reported by 36.2% of the sample (n = 690).

Dropout Rates and Reasons

Taking a lifetime service use perspective, 44.8% (n = 855) of men reported discontinuing therapy prematurely in the past without discussing this with their therapist. In other words, these men were considered to have dropped out of therapy. Next, recognizing that participants may have dropped out of multiple occasions of care in the past, the dropout rate was examined among men reporting only one previous occasion of care. Of this subgroup of 451 participants, 26.6% (n = 120) reported discontinuing prematurely without prior discussion with their therapist (i.e., dropping out).

Reasons for dropout were examined among participants who had dropped out of therapy without discussing this with their therapist (n = 855). The most common reason was a reported lack of connection or understanding between client and therapist (54.9%; n = 469), followed by the clients’ sense that therapy was unhelpful or “didn’t feel right” (20.2%; n = 173). The expense of therapy or logistical inconvenience was reported as the reason for dropout among 18.0% of participants (n = 154). Finally, a minority reported they dropped out because their issues were resolved (5.5%; n = 47), with the reason for dropout not specified or undetermined for 1.4% of the sample (n = 12).

Univariate Analyses Comparing Dropouts and Non-Dropouts

Results comparing past dropout men to men who did not report dropout are detailed in Table 1. Participants reporting past dropout were younger on average (mean difference = 3.96 years; d = 0.26), and dropout was significantly more likely among unemployed men (adjusted standardized residual = 3.6). Men who had dropped out of therapy reported lower scores on the TMF scale, indicating they identified themselves as less traditionally masculine (mean difference = 1.19; d = 0.19). Men who had dropped out also reported less optimism that therapy would help (d = 0.11) and less motivation to attend (d = 0.18) prior to commencing therapy, though effects were negligible to small. Men who had dropped out also reported greater feelings of emasculation in attending therapy (d = 0.16), and experienced greater shame (d = 0.12), upon commencing therapy, again small effects. Evident in the average number of therapist engagement strategies that clients reported occurred, stronger effects were observed for the level of engagement in therapy being lower among men reporting dropout (d = 0.23); with those reporting dropping out also experiencing elevated current depression symptom levels relative to non-dropouts (mean difference = 2.88; d = 0.43). Current suicide ideation was also more common among dropouts relative to non-dropouts (OR = 1.7).

Table 1.

Comparisons Between Male Client Participants Reporting Past Dropout or Not on Key Variables.

Variable	Group/Item		Dropout	No dropout	t/chi, p
			N = 855	N = 1052
Age	-	M (SD)	41.93 (14.50)	45.89 (15.54)	5.697, <.001
Sexuality	Heterosexual	% (n)	44.7 (615)	55.3 (761)	0.039, .843
	Non-heterosexual	% (n)	45.2 (240)	54.8 (291)
Location	Metropolitan	% (n)	45.8 (526)	54.2 (622)	3.80, .150
	Regional	% (n)	44.7 (277)	55.3 (342)
	Rural or remote	% (n)	37.1 (52)	62.9 (88)
Relationship	Single	% (n)	48.0 (244)	52.0 (264)	3.179, .204
	Partnered	% (n)	44.1 (491)	55.9 (623)
	Separated/widowed	% (n)	42.1 (118)	57.9 (162)
Employment	Employed (full-time, part-time, or casual)	% (n)	44.7 (562)	55.3 (695)	27.561, <.001
	Unemployed	% (n)	54.7 (150)	45.3 (124)
	Retired	% (n)	31.0 (65)	69.0 (145)
	Student	% (n)	47.0 (78)	53.0 (88)
Education	High school	% (n)	47.2 (222)	52.8 (248)	2.494, .476
	Trade/cert/diploma	% (n)	42.5 (227)	57.5 (307)
	Undergrad degree	% (n)	45.7 (227)	54.3 (270)
	Postgrad degree	% (n)	44.1 (179)	55.9 (227)
Traditional masculinity	TMF total score	M (SD)	28.04 (6.04)	29.23 (6.29)	4.201, <.001
Pre-therapy items	Perceived helpfulness[a]	M (SD)	3.07 (1.07)	3.19 (1.04)	2.452, .014
	Perceived motivation[b]	M (SD)	3.23 (1.29)	3.46 (1.25)	3.970, <.001
	Less of a man[c]	M (SD)	1.83 (1.17)	1.65 (1.06)	3.480, .001*
	Perceived shame[d]	M (SD)	2.13 (1.32)	1.98 (1.26)	2.437, .015*
Therapist engagement strategies		M (SD)	11.31 (2.72)	11.86 (2.05)	4.893, <.001*
Depression	PHQ-9 total score	M (SD)	11.59 (6.91)	8.71 (6.58)	9.260, <.001
Current suicidality	Yes	% (n)	53.2 (367)	46.8 (323)	30.505, <.001
	No	% (n)	40.1 (488)	59.9 (729)

Note. *Estimates adjusted for violation of the assumption of equal variance. aBefore you started therapy, to what extent did you think it would help? bBefore you started therapy, how motivated were you to attend? cDid going to therapy make you feel like less of a man? dTo what extent did you feel shame going to therapy? TMF = Traditional Masculinity Femininity Scale. PHQ = Patient Health Questionnaire. Bold = p < .05.

Multinomial Regression Analysis Predicting Dropout Reasons

A multinomial regression analysis was then conducted with dropout reasons relative to no dropout as the outcome variable. Independent variables were selected based on significance in the univariate tests conducted above: age, TMF total score, number of therapist engagement strategies, pre-therapy optimism, motivation, emasculation, and shame. Participants who did not report a reason for dropping out were excluded from this analysis, leaving a subsample of 1895 participants.

The overall model represented a good fit to the data, according to both a likelihood ratio test relative to an intercept-only model: χ2(40) = 132.46, p < .001; and a Pearson chi-square goodness-of-fit test: χ2(7512) = 7521.51, p = .467. The Nagelkerke R2 statistic for the overall model was .074. Overall likelihood ratio tests of individual predictors were significant for age: χ2(4) = 11.31, p = .023; number of therapist engagement strategies: χ2(4) = 26.62, p < .001; and employment: χ2(12) = 21.32, p = .046. Likelihood ratio tests were nonsignificant for TMF total score (p = .05); pre-therapy feelings that therapy would help (p = .069); motivation to attend (p = .126); feelings of emasculation in attending therapy (p = .163); and feelings of shame in attending therapy (p = .061).

Predictors of individual reasons for dropping out relative to no dropout are presented in Table 2. The odds of dropping out due to lack of connection with the therapist were greater with younger age (p = .02), unemployment relative to employment (p = .04), less identification with masculinity (p = .01), less evidence of therapist engagement (p < .001), and greater feelings of emasculation in attending therapy (p = .04). The odds of dropping out as a result of therapy “not working” were greater with unemployment relative to employment (p = .01), less evidence of therapist engagement (p = .001), less initial optimism that therapy would help (p = .01), and greater feelings of shame in attending therapy (p = .02). Greater odds of dropping out due to the expense or inconvenience of treatment were observed with younger age (p = .01). Finally, no significant predictors of dropping out due to resolved issues were identified (all p > .05; see Table 2).

Table 2.

Comparisons Between Male Client Participants Reporting Past Dropout or Not on Key Variables.

Dropout reason	Predictor		B	SE	OR	95% CI
“I didn’t connect with the therapist”n = 469	Age		−0.01	0.45	0.99	0.98	0.99
	Employment (employed)		-	-	-	-	-
		Unemployed	0.34	0.16	1.40	1.02	1.92
		Retired	−0.08	0.22	0.93	0.60	1.42
		Student	−0.33	0.22	0.72	0.47	1.12
	TMF total score		−0.03	0.01	0.97	0.96	0.99
	Therapist engagement		−0.11	0.02	0.90	0.86	0.94
	Perceived helpfulness[a]		0.05	0.07	1.05	0.92	1.19
	Perceived motivation[b]		−0.11	0.06	0.90	0.81	1.00
	Less of a man[c]		0.12	0.06	1.13	1.00	1.26
	Perceived shame[d]		−0.04	0.05	0.96	0.87	1.06
“Therapy didn’t work / didn’t feel right”n = 173	Age		−0.01	0.01	0.99	0.98	1.01
	Employment (employed)		-	-	-	-	-
		Unemployed	0.59	0.22	1.80	1.17	2.77
		Retired	−0.65	0.38	0.53	0.25	1.10
		Student	0.21	0.30	1.23	0.69	2.20
	TMF total score		0.00	0.01	1.00	0.98	1.03
	Therapist engagement		−0.11	0.03	0.90	0.85	0.96
	Perceived helpfulness[a]		−0.23	0.10	0.79	0.66	0.95
	Perceived motivation[b]		0.05	0.08	1.05	0.90	1.23
	Less of a man[c]		0.05	0.08	1.05	0.90	1.24
	Perceived shame[d]		0.16	0.07	1.17	1.02	1.34
“Therapy was too expensive or inconvenient”n = 154	Age		−0.02	0.01	0.98	0.97	0.99
	Employment (employed)		-	-	-	-	-
		Unemployed	0.22	0.25	1.24	0.76	2.03
		Retired	−0.34	0.40	0.71	0.33	1.56
		Student	−0.32	0.32	0.73	0.39	1.36
	TMF total score		−0.02	0.02	0.98	0.95	1.01
	Therapist engagement		−0.04	0.04	0.96	0.89	1.03
	Perceived helpfulness[a]		0.08	0.10	1.08	0.89	1.32
	Perceived motivation[b]		−0.15	0.09	0.86	0.73	1.02
	Less of a man[c]		0.17	0.09	1.18	1.00	1.40
	Perceived shame[d]		−0.02	0.08	0.98	0.84	1.14
“My issues were resolved”n = 47	Age		−0.01	0.01	0.99	0.96	1.01
	Employment (employed)
		Unemployed	0.11	0.43	1.12	0.48	2.59
		Retired	−0.88	0.79	0.42	0.09	1.96
		Student	−0.71	0.65	0.49	0.14	1.74
	TMF total score		−0.02	0.03	0.98	0.93	1.03
	Therapist engagement		−0.01	0.07	0.99	0.86	1.13
	Perceived helpfulness[a]		−0.01	0.17	0.99	0.71	1.38
	Perceived motivation[b]		−0.07	0.15	0.93	0.70	1.23
	Less of a man[c]		0.12	0.14	1.13	0.85	1.50
	Perceived shame[d]		0.18	0.12	1.19	0.94	1.52

Note. aBefore you started therapy, to what extent did you think it would help? bBefore you started therapy, how motivated were you to attend? cDid going to therapy make you feel like less of a man? dTo what extent did you feel shame going to therapy? TMF = Traditional Masculinity Femininity Scale. Bold = p < .05.

Discussion

Summary of Results

This study aimed to quantify the rates of, and reasons for, dropout from mental health talk therapy services among a large and diverse sample of Australian men, alongside modeling predictors of dropout according to various reasons identified by participating men. Considering lifetime service-use, results indicated 44.8% of respondents had dropped out of therapy without prior discussion with their therapist. For the subgroup of men who were reporting on their first and only experience in therapy (e.g., those having only accessed therapy once and not since returned), 26.6% reported having dropped out. Univariate comparisons identified younger age, unemployment, subjective identification with masculinity, pre-therapy feelings, therapist engagement strategies, and current depression all as factors significantly differentiating those who had dropped out in the past. Additionally, multivariable modeling of dropout reasons in relation to these predictors presented younger and unemployed men as more likely to have dropped out due to a lack of connection with their therapist, or due to therapy “not working” for unemployed men, and due to the expense of treatment for younger men. Less masculine-identifying men and those who felt emasculated in attending therapy were more likely to have dropped out due to a lack of connection with their therapist. Finally, men’s perception of less therapeutic engagement from their therapist predicted dropout due to both a lack of therapist–client connection and the feeling that therapy “didn’t work.” Dropping out due to “unhelpful” therapy was also predicted by lower pre-therapy optimism that treatment would help and greater shame in attending therapy among men. No significant predictors of dropping out due to resolved issues were identified.

Understanding Dropout: Rates and Reasons

Relative to previous studies documenting dropout in fixed cohorts with a single disorder in a controlled or clinical setting (e.g., Cooper & Conklin, 2015; Hans & Hiller, 2013), the online data captured here augment previous insights to increase awareness of men’s mental health therapy service experiences. Our results reflect men’s complex trajectories in and out of services depending on a variety of factors. Appraising dropout itself as universally problematic and always avoidable has recently been challenged (Leichsenring et al., 2019). This is because for some men (and clients in general), dropping out of therapy may not necessarily reflect a negative outcome, but more so a process of weighing up the relative cost and necessity of engagement with a therapy service at a given time, and as a bridge to effective self-management. Individuals also commonly see a number of therapists prior to finding one with whom an effective working alliance can be achieved (Chen et al., 2017). This is often thought to represent a “try before you buy” mentality (O’Keeffe et al., 2019), particularly as it is known that men’s preferences for types and styles of therapy can vary substantially (Kealy et al., 2020; Zajac et al., 2020). This could explain the ostensibly high “lifetime” service use dropout rate observed. Studies examining dropout have not conceptualized men’s dropout as a “revolving door” phenomenon, where men fluctuate between engagement and disengagement with services over time. In promoting the experience of men in this study and respecting their experiential knowledge, the present findings validate the importance of privileging shared decision-making through a person-centered approach over more reductionist, paternalistic views of therapy dropout as “failure,” which assumes that male clients are unaware of the severity of their mental health concerns (Leichsenring et al., 2019).

It is essential to also reflect on the findings presented here in the context of past work showing that a dissatisfying therapy experience for men can increase doubt and mistrust in services and act as a barrier to further re-engagement (Richards & Bedi, 2015; Seidler, Rice, Kealy et al., 2020). The fact that more than one in four participants experienced dropout following their first therapy experience and had not re-engaged with a service at the time of responding, is of concern. Univariate analyses also highlighted that rates of current suicidality were elevated among men who had dropped out of their first and only experience. Previous research notes that at-risk men experiencing suicidality are even more reluctant to take up therapy (Cleary, 2017; Harris et al., 2016; River, 2018) and that previous dissatisfaction with therapy directly affects a man’s willingness to later disclose distress to their family physician, the gatekeepers for public system help-seekers in Australia (Seidler, Rice, Kealy et al., 2020). Taken together, this evidence highlights the importance of ensuring that when men with complex needs do reach out, therapists are prepared to investigate and meet their needs, early and effectively, to fundamentally shift the narrative away from homogenizing men as universally averse to help-seeking (River, 2018).

The most commonly reported reason for dropout among the sample was a lack of connection or understanding between men and their therapist. This reinforces existing findings that problems in the therapeutic relationship and discordant client–therapist expectations account for the most variance in dropout (Garcia & Weisz, 2002; Swift & Greenberg, 2012). Interestingly, less than 6% of men reported dropping out prematurely because their issues were resolved, and no significant predictors of dropping out due to alleviated issues were identified. Whilst this may be explained by low numbers of men reporting dropout due to resolved problems, this finding nevertheless challenges existing assumptions based on therapist-led findings that many dropouts may stem from an uncommunicated understanding that therapy is deemed complete (O’Keeffe et al., 2019; Warnick et al., 2012). Rather, our findings indicate that therapy dropout occurs most often because of an unhelpful, dissatisfying process, characterized by a lack of shared understanding. Thus, the findings illustrate the stark reality of the need to better target mental health services through a person-centered, gender-sensitized approach for men (River, 2018; Seidler, Rice, River, et al., 2018). This is reinforced by the fact that one-fifth of men who had experienced dropout reported they decided to leave therapy due to lack of progress and the discomfort associated with letting one’s guard down in an unfamiliar environment. These men’s experiences suggest that some therapists are either not adequately recognizing or responding to men’s needs.

Sociodemographic Predictors of Dropout

Past experience of dropout due to a lack of connection between client and therapist was more likely among younger men and men experiencing unemployment—findings that extend on patterns observed in previous studies (Edlund et al., 2002; Henzen et al., 2016; Seidler, Rice, Dhillon, et al., 2020). This may reflect the inherent challenge tied to mental health service engagement among men from these demographics, as in many cases, younger men and men experiencing social disadvantage do not necessarily have the resources or social scaffolding that would allow them to continue with their therapy (Rice et al., 2018). Intersectional understanding of the overlap between gender and health has long understood younger men as identity forming and unemployed men as experiencing health inequities, both of which appear to intersect with masculinities to deter mental health service engagement among these men (Griffith et al., 2016). These subgroups are uniquely at risk of suicide (King et al., 2020; Milner et al., 2014), highlighting a unique and complex interaction between masculine identity formation, mental health stigma, and help-seeking among young and/or unemployed men requiring further investigation. Given unemployed men were more likely to drop out due to the feeling that therapy “didn’t work” and younger men were more likely to drop out due to the expense of treatment, perhaps men from these demographics require a higher level of proof that therapy is worth their time given their social situations may not be most conducive to engagement with often expensive and time-consuming therapy.

Contrary to our hypotheses and past literature, education level was unrelated to risk of dropout (Egan & Kenny, 2005). Unexpected was the finding of no association between participants’ location and dropout risk. Rural and regionally located men have been found identified to be at greater risk of service disengagement in past studies, due to additional barriers and stigma surrounding effective service access (Seidler, Rice, Dhillon, et al., 2020). While encouraging that rural men did not experience greater risk of dropout, this finding requires further clarity and replication.

Therapist Engagement and Dropout

Dropping out due to a lack of therapist–client connection and the sense that therapy “didn’t work” for men were both more likely when men felt their therapist expended less effort to engage them. These findings suggest that men may be reluctant to trust or invest in a process that assumes rather than asks when it comes to ways of working in therapy. This reinforces the need for the client’s expectations and beliefs around help-seeking to be appropriately addressed through a purposeful orientation and education to the service, to establish a collaborative course of care. In order to capitalize on the short window of opportunity that exists when engaging men (Seidler, Rice, Ogrodniczuk, et al., 2018), this approach must recognize the man’s distress as it presents and his unique journey in overcoming a plethora of barriers to attend (Richards & Bedi, 2015). Men who were less hopeful that therapy would help, and who felt greater shame in attending, were also more likely to have dropped out due to feeling that therapy was unhelpful—highlighting a clear need for therapists to enact targeted engagement to overcome these psychological barriers to engagement. For example, focusing on increasing self-compassion as a means to reduce feelings of shame and self-stigma surrounding help-seeking has been found useful, albeit it is important to consider how this is best framed to be acceptable for male clients (Heath et al., 2017). Addressing this shame is particularly pertinent considering rates of suicidality were significantly elevated among men who had experienced dropout. Previous work with suicidal men highlights a common rejection of services that frame emotional distress as mental illness over contextualizing issues as situational or relational stressors (River, 2018). Further complicating clinical work with men is that many may express their internalized distress through more socially condoned, traditionally masculine “externalizing” symptoms of anger, aggression, or substance use, which therapists may interpret as normative masculinities or feel either unequipped or uncomfortable responding to (Brownhill et al., 2005; Martin et al., 2013; Rice et al., 2013). Continuing to adopt this strategy to the provision of mental health services for men may help to alleviate the extent to which dropout occurs unnecessarily, due to failure to provide purposeful engagement and subsequent therapy that is sensitized to men’s needs.

Masculinity and Dropout

Participants reporting a sense of emasculation (i.e., that presenting to therapy made them feel like “less of a man”) while in therapy experienced higher rates of dropout due to a lack of connection with their therapist. This novel finding underscores the importance of therapists’ capacity to enact “gender competency” when working with men. Gender competency is a novel construct reflecting the capacity to effectively implement male-specific adaptations to therapy, alongside accommodating men’s unique experiences of mental ill health as a potential extension of their experience of masculinity (Owen et al., 2009). Without appropriate awareness and training, therapists may be unaware of their beliefs and biases about men’s mental health and become complicit in reinforcing rigid and unhelpful stereotypes about male emotionality or help-seeking (Seidler, Rice, Dhillon, et al., 2019). For example, Almaliah-Rauscher et al. (2020) highlighted that therapists are often less confident and unwilling to treat male clients with complex needs such as suicidality than they are female clients—a bias that may problematize the extent to which therapists can provide an effective service, or referral pathway for men, potentially explaining the observed association between lack of therapy engagement and dropout. Conversely, men identifying more strongly with traditional masculinity were less likely to have dropped out due to a lack of therapist–client connection. Considered in relation to the finding that feeling emasculated in therapy predicted greater odds of dropout, perhaps this indicates that regardless of the extent to which men identify with traditional masculinities, feeling emasculated can have a negative impact on men’s willingness to engage with a course of therapy. This finding may also reflect the capacity for masculinity norms to enact positive influences on mental health (Levant & Wimer, 2014), whereby men may display stoicism in their pursuit of positive outcomes in therapy. This challenges the prevailing narrative in the literature that therapy is unapproachable for men who endorse traditional masculinities. Working alongside male socialization in therapy represents a complex process with opportunities for success if traditional masculinities are approached purposefully and leveraged effectively as a vehicle toward therapeutic change.

Limitations and Future Directions

The current study has several limitations. First, the design of the survey was such that men’s experiences were retrospectively assessed through cross-sectional self-report. The results may have been impacted by recall bias, particularly for participants whose most recent experience of therapy occurred over 5 years ago. Second, the small magnitude of between-group differences when exploring dropout predictors must be considered, with the size of the sample making statistical significance likely, while clinical significance remains harder to quantify. The naturalistic selection of diverse male participants in this study rather than the more common controlled therapy setting gives these results a unique external validity in the field. The reported rates and any group-based findings should not be considered in isolation or as a broad indictment of the mental health therapy, but more so as the continuation of a nuanced narrative galvanizing a call to action for therapists to consider how they work with certain men. Third, the current findings were also limited by the data available, as many other factors known to affect mental health service use, beyond the scope of the current study, were unavailable for analysis. These include the type of service or therapists accessed, length of therapy, wait times, and medication use, in addition to potentially relevant patient characteristics including stigma, symptom severity at point of entry and dropout, and any functional impairments. Finally, while the “experiential knowledge” of male clients was sought and privileged here, the treating therapists’ perspectives on working with the male respondents and their specific characteristics (e.g., level of training, communication skills, and clinical expertise) were not examined.

To address some of these limitations, the field will benefit from conducting in-depth qualitative studies among men who have experienced dropout both recurrently over their lives and following only one dissatisfying occasion of care. Including the treating therapist and their experience in these discussions where possible will help to further our understanding of dropout and the contribution of both client and therapist factors to this process, including the extent to which dropout is not necessarily a negative outcome. Practically, on the basis of an amassed understanding of the nature of and reasons for dropout among men, clinical communities will benefit from co-designed, targeted engagement protocols for therapists working with vulnerable groups of men. As a preliminary recommendation, the findings here underscore a need for therapists to adopt an open and honest conversation about therapy ambivalence, resistance, and dropout during their initial interactions with a client, to ensure that the process of dropping out and re-engaging with subsequent therapists is appropriately normalized and emblematic of the need for clients to achieve an effective working alliance, regardless of therapist. This may help buffer against the extent to which dropout manifests as a deterrent to future service access.

Conclusion

The benchmarking of rates of dropout achieved here represents a necessary shift toward appraising the success of our mental health services in engaging men from a within-men perspective. By reporting on men’s perspectives and examining men’s experiences of therapy engagement, findings from the present study can advance our understanding of therapy dropout among men. Whilst clearly the present rates of dropout require further delineation and more fine-grained quantification in future work, this reporting represents a crucial initial exploration of the rates and predictors of dropout from mental health therapy among men in the community, highlighting the need to focus on young and unemployed men in particular—a level of understanding not achieved in the literature thus far.

Click here for additional data file.

Supplemental material, sj-pdf-1-jmh-10.1177_15579883211014776 for Men’s Dropout From Mental Health Services: Results From a Survey of Australian Men Across the Life Span by Zac E. Seidler, Michael J. Wilson, David Kealy, John L. Oliffe, John S. Ogrodniczuk and Simon M. Rice in American Journal of Men’s Health