Motivational interviewing to enhance treatment attendance in mental health settings: A systematic review and meta‐analysis

Probably not a substantial claim: may not pass S3 (load-bearing).

In Maltby and Tolin, the sample had high motivation, with 57% in the action or maintenance stage of change.

“The sample comprised participants with high levels of motivation, where 57% claimed their stage of change category being either in the action or maintenance phase.”

May combine two claims.

The included studies had restricted methodological quality, with biases mainly due to inconsistent reporting of blinding of sample, researchers, and clinicians.

“Methodological quality of the included studies was restricted, and biases of the studies were due mainly to blinding issues, where blinding of the sample, researchers and clinicians was not reported consistently.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Two included studies self-reported possible sample bias due to recruitment methods.

“Two studies self-reported their sample may be biased due to recruitment methods (Buckner & Schmidt, 2009;Maltby & Tolin, 2005).”

Classified as probably not a substantial claim.

Fourteen randomized controlled trials were included in the full review.

“Fourteen RCTs were identified for inclusion in the full review, and are summarized in Table 1.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Four included studies reported small sample size as a limitation, lacking power to detect effects on study outcomes.

“Four studies reported small sample size as a limitation, lacking power to detect effects on study outcomes (Buckner & Schmidt, 2009;Martino et al., 2006;Simpson et al., 2010).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

A majority of included studies reported unequal gender recruitment, but recruitment settings and sample types reflected the disproportionate sampling.

“A majority of the studies reported unequal gender recruitment, but the settings for recruitment and sample type reflected the disproportionate sampling.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Three studies recruiting participants with anxiety disorders or sensitivity reported higher female participation, and four studies recruiting for dual diagnosis reported larger male participation.

“Three studies that recruited participants with anxiety disorders or sensitivity (Buckner & Schmidt, 2009;Korte & Schmidt, 2015;Westra & Dozois, 2006) reported a higher female participation (63%, 91% and 70%, respectively), and four studies recruiting for dual diagnosis reported a larger male participation (62-75%) (Baker et al., 2002;Martino et al., 2000Martino et al., , 2006;;Swanson et al., 1999”

May combine two claims.

The 14 trials included 803 participants, with a mean sample size of 57.4 and a range of 12 to 160.

“A total of 803 participants were included in the 14 trials with a mean sample size of 57.4 (range 12-160) (refer to Table 1).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Eight studies recruited non-treatment-seeking samples, and six recruited treatment-seeking samples.

“Eight studies recruited non-treatment-seeking samples (Buckner & Schmidt, 2009;Fiszdon et al., 2016;Maltby & Tolin, 2005;Seal et al., 2012;Zanjani et al., 2008).The remainder (n = 6) recruited treatment-seeking samples (Baker et al., 2002;Korte & Schmidt, 2015;Martino et al., 2000Martino et al., , 2006;;Simpson et al., 2010;Swanson et al., 1999;Syzdek et al., 2014Syzdek et al., , 2016;;Westra & Do”

• cites Buckner & Schmidt, 2009Not checked yet

  Not checked in the free run.

• cites Fiszdon et al., 2016Not checked yet

  Not checked in the free run.

• cites Maltby & Tolin, 2005Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Only one included study conducted follow-up by checking medical records for attendance at 4, 8, and 16 weeks.

“Only one study (Seal et al., 2012) conducted follow-up by checking medical records for evidence of attendance (at 4, 8 and 16 weeks)”

Probably not a substantial claim: may not pass S3 (load-bearing).

The approximate gender distribution across the entire sample was 546 males and 257 females.

“The approximate gender distribution for the entire sample was 546 (68%) males and 257 (32%) females.”

May combine two claims.

All included studies reported attrition rates and accounted for all patients at study conclusion.

“All studies reported attrition rates and accounted for all patients at the study conclusion.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Most included studies were conducted in outpatient settings, with two exceptions (Baker et al., 2002; Swanson et al., 1999), and one study recruited both inpatients and outpatients (Martino et al., 2006).

“With the exception of two studies (Baker et al., 2002;Swanson et al., 1999), most studies were conducted in outpatient settings.Martino et al. (2006) recruited both inpatients and outpatients.”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Swanson et al., 1999Not checked yet

  Not checked in the free run.

• cites Martino et al., 2006Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

The number and duration of MI interventions varied widely, from 1-2 phone calls of 15 minutes each to three face-to-face sessions totaling 6.5 hours.

“The number and duration of MI interventions varied, ranging from 1 to 2 phone calls for 15 min each (total 30 min) (Zanjani et al., 2008) to three face-to-face sessions totalling 6.5 hr (Buckner & Schmidt, 2009) (see Table 3).”

• cites Zanjani et al., 2008Not checked yet

  Not checked in the free run.

• cites Buckner & Schmidt, 2009Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Buckner and Schmidt (2009) reported therapists' mean global rating scale ranged from 6.11 to 7.00 (mean 6.45, SD 0.72), indicating competence for MI (above 6 recommended).

“Buckner and Schmidt (2009) reported that the therapists' mean global rating scale ranged from 6.11 to 7.00 (mean 6.45, SD 0.72), and were competent for MI (a rating above 6 was recommended).”

• cites Buckner and Schmidt, 2009Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

All studies except Seal et al. (2012) described a script or protocol for the MI intervention.

“All studies, with the exception of Seal et al. (2012), described a script or protocol for the MI intervention.”

• cites Seal et al., 2012Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Several tools, such as the MITI coding, were used to report MI fidelity, but studies varied in the level of detail provided.

“Several tools, such the Motivational Interviewing Treatment Integrity (MITI) coding (Moyers et al., 2010), were used to report MI fidelity (see Table 5), but studies varied in the level of detail provided.”

• cites Moyers et al., 2010Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Five studies included individuals with subthreshold symptoms of mental illness.

“Five studies had criteria that included individuals with subthreshold symptoms of mental illness.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Fiszdon et al. (2016) formally evaluated a random sample of 20% of recorded interviews.

“However, Fiszdon et al. (2016) formally evaluated a random sample of 20% of recorded interviews (see Table 4).”

• cites Fiszdon et al., 2016Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Martino et al. (2006) and Swanson et al. (1999) indicated absence of a nontreatment group was a limitation, and Westra and Dozois (2006) indicated same therapist for both groups was a limitation.

“Martino et al. (2006) and Swanson et al. (1999) both indicated, however, that the absence of a nontreatment group was a limitation, while Westra and Dozois (2006) indicated that their study was limited by having the same therapist for both the MI group and comparison group.”

• cites Martino et al., 2006Not checked yet

  Not checked in the free run.

• cites Swanson et al., 1999Not checked yet

  Not checked in the free run.

• cites Westra and Dozois, 2006Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Seal et al. (2012) stated that 88% of statements during interviews were congruent with MI principles.

“Seal et al. (2012) stated that 88% of statements made during the interviews were congruent with MI principles.”

• cites Seal et al., 2012Not checked yet

  Not checked in the free run.

May combine two claims.

Only Simpson et al. (2010) reported MITI global scores and subscale scores for Evocation, Collaboration, Autonomy, and Direction; the Direction subscale was not MI congruent and was similar to control group scores.

“Only Simpson et al. (2010) reported the MITI ratings for global scores, and the subscale scores for Evocation, Collaboration, Autonomy and Direction.They specifically reported that the Direction subscale was not MI congruent and was similar to the control group scores.”

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

Fiszdon et al. (2016) found MI interviews had higher MI strategy adherence (6.05 vs 2.58, p=.001) and MI competence (4.48 vs 3.66, p=.006) than control, with no significant differences in general interview adherence or competence.

“Fiszdon et al. (2016) used a specially designed assessment form for their study and stated that the MI interviews were higher than the control group in regard to MI strategy adherence (6.05 vs. 2.58, p = .001)and MI competence (4.48 vs. 3.66, p = .006).There were no significant differences between the MI and control groups regarding general interview adherence or competence (p > .05).”

• cites Fiszdon et al., 2016Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In five studies, comparison interventions were no treatment or minimal treatment, while Seal et al. (2012) used four short phone calls over 8 weeks to discuss appointment logistics.

“In five studies, the comparison interventions were either no treatment or minimal treatment (Maltby & Tolin, 2005;Syzdek et al., 2014Syzdek et al., , 2016;;Westra & Dozois, 2006;Zanjani et al., 2008), while Seal et al. (2012) conducted four short phone calls over 8 weeks to discuss logistics regarding appointments.”

• cites Maltby & Tolin, 2005Not checked yet

  Not checked in the free run.

• cites Syzdek et al., 2014Not checked yet

  Not checked in the free run.

• cites Westra & Dozois, 2006Not checked yet

  Not checked in the free run.

May combine two claims.

Half the studies did not mention whether therapists were exclusive to MI or conducted both interventions, and only three studies had therapists exclusive to MI.

“Half the studies did not mention whether or not study therapists were exclusive to the MI intervention or conducted both interventions (Baker et al., 2002;Buckner & Schmidt, 2009;Fiszdon et al., 2016;Maltby & Tolin, 2005;Seal et al., 2012;Swanson et al., 1999;Syzdek et al., 2014), and only three studies had study therapists that were exclusive to the MI intervention (Syzdek et al., 2016;Westra &”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Buckner & Schmidt, 2009Not checked yet

  Not checked in the free run.

• cites Fiszdon et al., 2016Not checked yet

  Not checked in the free run.

May combine two claims.

Baker et al. (2002) found no difference in attendance to offered treatment (16% vs 17.3%), and the control group attended more post-to pre-treatment sessions than the MI group (5.79 vs 4.46).

“The study by Baker et al. (2002) also found no difference in attendance to the offered treatment (16% vs. 17.3%), and reported that the control group attended a greater number of post-to pre-treatment sessions compared to the MI pretreatment group (5.79 vs. 4.46, respectively).”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

May combine two claims.

Five studies reported minimal or no effect of MI as a pre-treatment, including three treatment-seeking and two non-treatment-seeking samples.

“Five studies reported minimal or no effect of MI as a pre-treatment, of which three recruited treatment-seeking participants (Baker et al., 2002;Martino et al., 2006;Simpson et al., 2010), and two recruited non-treatment-seeking samples (Syzdek et al., 2014(Syzdek et al., , 2016)).”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Martino et al., 2006Not checked yet

  Not checked in the free run.

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

Visual inspection of heterogeneity showed Baker et al. (2002) and Simpson et al. (2010) did not intersect, and both reported no differences between groups.

“Heterogeneity assessment by visual inspection of the vertical line from mid-points of the black diamond shows that Baker et al. (2002) and Simpson et al. (2010) did not intersect.Both studies reported no differences between intervention and control groups.”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

Syzdek et al. (2016) found a significant increase at two-month follow-up for the MI group to seek informal help from a parent (45% vs 8%), and a nonsignificant trend for professional sources (39% vs 8%).

“Syzdek et al. (2016) found there was a significant increase at two-month follow-up for the MI group to seek informal help from a parent (45% vs. 8%), and a nonsignificant trend for the MI group to seek help from professional sources (39% vs. 8%).”

• cites Syzdek et al., 2016Not checked yet

  Not checked in the free run.

May combine two claims.

Martino et al. (2006) found that despite more MI participants attending more post-MI interventions (75% vs 55%), they attended fewer sessions and had reduced attendance; no differences in mean days (19.16 vs 19.09, p=.66) and no participant remained at 12 weeks.

“Martino et al. (2006) found that despite more participants from the MI pretreatment intervention attending more post-MI interventions (75% vs. 55%), they attended fewer sessions in the offered programme and had reduced attendance.There were no differences between the MI intervention and standard treatment mean days (19.16 vs. 19.09,respectively, p = .66)and no participant from either group”

• cites Martino et al., 2006Not checked yet

  Not checked in the free run.

May combine two claims.

Simpson et al. (2010) found no significant difference between MI pre-treatment and standard care in attending or completing post-MI treatments (all randomized p=.23; completers p=.13).

“Simpson et al. (2010) found no significant difference between MI pre-treatment intervention and standard care groups in attending or completing post-MI treatments (all randomized: p = .23;all completers: p = .13).”

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

In the treatment-seeking subgroup, the intervention group was more likely to attend treatment (OR: 1.79, 95% CI: 0.81-3.96) but the effect was not significant (Z = 1.44, p = .15).

“The RE model revealed that although the intervention group was more likely to attend treatment (OR: 1.79, 95% CI: 0.81-3.96), the effect was not significant (Z = 1.44, p = .15).”

May combine two claims.

The treatment-seeking subgroup included five studies with 410 participants (207 intervention, 203 control).

“The treatment-seeking group included five studies (Baker et al., 2002;Martino et al., 2006;Simpson et al., 2010;Swanson et al., 1999;Westra & Dozois, 2006), with an overall sample of 410 participants (207 intervention group, 203 control group).”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Martino et al., 2006Not checked yet

  Not checked in the free run.

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

In the treatment-seeking subgroup, heterogeneity was substantial (I2 = 57%) and all confidence intervals intersected the line of no effect.

“Heterogeneity was substantial (I2 = 57%), and all confidence intervals intersected the line of no effect.”

May combine two claims.

Within the treatment-seeking sample, Simpson et al. (2010) reported an OR of 0.20 (95% CI: 0.02-2.02) and Baker et al. (2002) reported an OR of 0.94 (95% CI: 0.41-2.16), suggesting no effect for MI.

“Within this sample, Simpson et al. (2010) reported an OR of 0.20 (95% CI: 0.02-2.02),and Baker et al. (2002) reported an OR of 0.94 (95% CI: 0.41-2.16),suggesting no effect for the MI intervention (refer to Figure 2).”

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

May combine two claims.

The non-treatment-seeking subgroup included seven studies with 301 participants (152 intervention, 149 controls).

“Seven studies were included in the non-treatment-seeking analysis component (Buckner & Schmidt, 2009;Korte & Schmidt, 2015;Maltby & Tolin, 2005;Seal et al., 2012;Syzdek et al., 2014Syzdek et al., , 2016;;Zanjani et al., 2008), with a smaller overall sample size (n = 301 participants; 152 in the intervention group and 149 controls).”

• cites Buckner & Schmidt, 2009Not checked yet

  Not checked in the free run.

• cites Korte & Schmidt, 2015Not checked yet

  Not checked in the free run.

• cites Maltby & Tolin, 2005Not checked yet

  Not checked in the free run.

May combine two claims.

In a sensitivity analysis removing Baker et al. (2002) and Simpson et al. (2010), the sample was homogenous (I2 = 0%) and participants who attended MI pre-treatment were more likely to seek post-MI treatment (OR: 4.04, 95% CI: 2.71-6.04, Z = 6.83, p < .001).

“A sensitivity analysis was conducted and the data from Baker et al. (2002) and Simpson et al. (2010) was removed from a re-run of the analysis.The sample was homogenous (I2 = 0%), and participants who attended MI pre-treatment were more likely to seek post-MI treatment (OR: 4.04, 95% CI: 2.71-6.04),with a significant effect size (Z = 6.83, p < .001).”

• cites Baker et al., 2002Not checked yet

  Not checked in the free run.

• cites Simpson et al., 2010Not checked yet

  Not checked in the free run.

May combine two claims.

In the non-treatment-seeking subgroup, all confidence intervals intersected the line of no effect, there was no heterogeneity (I2 = 0%), and the intervention group was more likely to attend treatment (OR: 4.83, 95% CI: 2.84-8.24, Z = 5.79, p < .001).

“All confidence intervals intersected the line of no effect, and there was no heterogeneity (I2 = 0%) and the RE model revealed that the intervention group was more likely to attend treatment (OR: 4.83, 95% CI: 2.84-8.24),with a large treatment effect size (Z = 5.79, p < .001)(see Figure 2).”

May combine two claims.

The funnel plot was symmetrical, indicating no evidence of publication bias.

“Subjectively, as the funnel plot is symmetrical, there is no evidence of publication bias (see Figure 3).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Motivation style counselling is also shown to be beneficial in this review.

“Otherwise, it would be more equivalent to a motivation style counselling, which is also shown to be beneficial in this review.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In that prior meta-analysis, motivational interviewing had no significant effect for eating disorder samples and a medium effect for anxiety, mood, and psychotic disorder samples.

“Subgroup analyses (eating disorders, mood, anxiety, psychotic) revealed MI achieved a nonsignificant effect for samples with eating disorders (d = 0.08), and a medium effect for samples with anxiety, mood and psychotic disorders (d = 0.54, p = .003).”

• cites Romano & Peters, 2015Not checked yet

  Not checked in the free run.

May combine two claims.

A prior meta-analysis found that pooled attendance was significantly higher for motivational interviewing samples, with a small-to-medium effect and substantial heterogeneity.

“Their pooled data demonstrated a significantly enhanced attendance for MI samples (d = 0.38, p = .012)but with substantial heterogeneity (I2 = 65.85,p < .001).”

• cites Romano & Peters, 2015Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

In Buckner and Schmidt, masking the study intention as an interview study of anxiety may have attracted already motivated participants.

“In the study by Buckner and Schmidt (2009), the researchers masked the study intention by describing it as an “interview study of anxiety,” which may have attracted already motivated participants to discuss and change behaviour.”

May combine two claims.

Only one prior meta-analysis of motivational interviewing and attendance had been conducted, and it showed substantial heterogeneity and did not specifically investigate non-treatment-seeking samples.

“However, only one meta-analysis appears to have been undertaken (Romano & Peters, 2015), which demonstrated substantial heterogeneity, and did not specifically investigate the effects of MI on non-treatment-seeking samples.”

• cites Romano & Peters, 2015Not checked yet

  Not checked in the free run.

May combine two claims.

A recent systematic review and meta-analysis of motivational interviewing mechanisms in psychotic, mood, anxiety, eating and comorbid conditions found that most studies reported few mechanisms of change and limited evidence for causal links to outcomes.

“The reviewers found that a majority of studies reported few MI mechanisms of change and there was also limited evidence for causal links to outcomes (Romano & Peters, 2015).”

• cites Romano & Peters, 2015Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In the stages of change model, individuals already seeking treatment are considered motivated and in the preparation or action stage, which is the desired outcome of successful motivational interviewing.

“In the stages of change model, individuals that are already seeking treatment would be considered to be motivated and in the “preparation” or the “action” stage of change, which is the desired outcome of successful MI interventions.”

• cites Prochaska, DiClemente, & Norcross, 1992Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

Little was known about the effectiveness of motivational interviewing as a pretreatment for both treatment-seeking and non-treatment-seeking individuals.

“Currently, little is known about the effectiveness of MI as a pretreatment for both (mental health) treatment-seeking and non-treatment-seeking individuals.”

May combine two claims.

Motivational interviewing has been used successfully with psychotic disorders, showing effectiveness in reducing excessive drinking.

“It has been used successfully with psychotic disorders by demonstrating effectiveness in the reduction of excessive drinking (Baker et al., 2012).”

• cites Baker et al., 2012Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Macdonald et al. found motivational interviewing results promising but difficult to compare because of heterogeneity between studies.

“Macdonald et al. (2012) found that the results were promising but difficult to compare due to heterogeneity between studies.”

• cites Macdonald et al. (2012)Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim. May combine two claims.

Dray and Wade found that causal factors for motivational interviewing effects were weak and that results were insufficient to assess motivational interviewing efficacy.

“Dray and Wade (2012), who also reviewed the literature regarding samples with eating disorders, found that, similarly to Romano and Peters (2015), the casual factors for the effects of MI were weak, and the results were insufficient to properly assess the efficacy of MI.”

• cites Dray and Wade (2012)Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Macdonald et al. found motivational interviewing most beneficial for increasing readiness to change and potentially useful for preparing individuals not ready to instigate change themselves.

“However, the results indicated that MI was most beneficial in regard to increasing “readiness to change” and may be useful in preparing individuals for change when they are not ready to instigate the change themselves.”

• cites Macdonald et al. (2012)Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Full texts were reviewed independently by two members of the review team against the inclusion criteria, with disagreements arbitrated by a third member.

“Full texts of these studies were then reviewed independently by two members of the review team against the inclusion criteria.Disagreements were arbitrated by a third team member.”

Probably not a substantial claim: may not pass S3 (load-bearing).

A funnel plot was generated to investigate potential reporting bias among the studies.

“A funnel plot was also generated to investigate potential reporting bias of the studies.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The observed gender pattern is consistent with previous studies indicating that women have a higher prevalence of mental illness while men have a higher prevalence of substance use and behavioural disorders.

“This is consistent with previous studies across all age groups that indicate that despite women having a higher prevalence of mental illness, men have a higher prevalence of substance use and behavioural disorders (AIHW, 2015).”

• cites AIHW, 2015Not checked yet

  Not checked in the free run.

May combine two claims.

Hettema et al. (2005) found that a manual-based protocol was the only factor predicting outcome (8.5% of variance), and studies without a manual reported higher effect scores (d=0.65) than those with one (d=0.37).

“Hettema et al. (2005) found that a manual-based protocol was the only associated factor that predicted outcome (8.5% of the variance), and studies that did not use a manual reported higher effect scores (d = 0.65) than those that had used one (d = 0.37).”

• cites Hettema et al., 2005Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Seal et al. (2012) based their intervention on findings from their pilot study and a meta-analysis indicating that MI effect size was not predicted by duration, purity, counsellor training, or post-training support.

“Seal et al. (2012) based their intervention on findings from their pilot study, and also from results of a meta-analysis (Hettema, Steele, & Miller, 2005) that indicated MI intervention effect size was not predicted by MI duration, purity, counsellor training or post-training support.”

• cites Hettema, Steele, & Miller, 2005Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

This finding is comparable to Hettema et al. (2005), which concluded MI may be contraindicated for individuals ready for change because they are already at a high level of motivation and in the preparation or action stage.

“This is comparable to a meta-analysis by Hettema et al. (2005), which concluded that MI may be contraindicated for individuals ready for change due to them already being at a high level of motivation and within the preparation or the action stage of the change cycle.”

• cites Hettema et al., 2005Not checked yet

  Not checked in the free run.

May combine two claims.

The non-treatment-seeking intervention group was homogenous and almost five times more likely to attend post-MI treatment than treatment-seeking samples.

“The nontreatment-seeking intervention group was homogenous and almost five times more likely to attend post-MI treatment than samples that were treatment seeking.”

May combine two claims.

An individual's perceived need is the strongest predictor of the use of mental health services, and the more severe the mental illness, the higher the perceived need.

“An individual's perceived need is the strongest predictor of the use of mental health services (Mills et al., 2012) and the more severe the mental illness, the higher was the perceived need (Andrade et al., 2014).”

• cites Mills et al., 2012Not checked yet

  Not checked in the free run.

• cites Andrade et al., 2014Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Public and personal stigma are known barriers to seeking treatment in some studies.

“Public and personal stigma are known barriers to seeking treatment in some studies (Corrigan et al., 2006).”

• cites Corrigan et al., 2006Not checked yet

  Not checked in the free run.

May combine two claims.

An individual's attitude is an important barrier to initiating and engaging in treatment, and attitudinal barriers are highly prevalent in mild and moderate cases of mental illness.

“An individual's attitude is an important barrier to initiating and engaging in treatment, where attitudinal barriers are highly prevalent in mild and moderate cases of mental illness (Andrade et al., 2014).”

• cites Andrade et al., 2014Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim. May combine two claims.

The rationale behind MI is that it raises an individual's ambivalence, taking into account their stage of change, to move them to the next stage by enhancing perceived need for change and increasing motivation to change.

“The rationale behind MI is that it raises an individual's ambivalence, taking into account their stage of change, to move that individual to the next stage by enhancing their perceived need for change, and increasing their motivation to change (Prochaska et al., 1992).”

• cites Prochaska et al., 1992Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim. May combine two claims.

Those not seeking treatment may be at the precontemplation or contemplation stage, and MI can assist them to increase motivation to change by exploring ambivalence.

“Those not seeking treatment may be at the precontemplation or contemplation stage, and MI can assist them to increase their motivation to change by exploring ambivalence.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Telephone interventions are relatively low cost and contextually appropriate for healthcare settings, particularly when fiscal considerations are paramount.

“are a relatively low cost and a contextually appropriate tool for use in healthcare settings, particularly when fiscal considerations are paramount (Kaplan, 2006)”

• cites Kaplan, 2006Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Other studies have demonstrated the efficacy of MI as a telephone-delivered intervention.

“other studies have also demonstrated the efficacy of MI as a telephone-delivered intervention (Gaume et al., 2014;Mello et al., 2008Mello et al., , 2012)).”

• cites Gaume et al., 2014Not checked yet

  Not checked in the free run.

• cites Mello et al., 2008Not checked yet

  Not checked in the free run.

• cites Mello et al., 2012Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Therapeutic alliance in telephone-delivered psychological therapies was comparable to face-to-face.

“therapeutic alliance was comparable”

• cites Jenkins-Guarnieri et al., 2015Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Telephone interventions do not have the limitations of computer-based interventions, which may be restricted due to Internet access and its unreliability in remote areas.

“and do not have the limitations of computer-based interventions, which may be restricted due to Internet access and its unreliability in remote areas (Harrison et al., 2011).”

• cites Harrison et al., 2011Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Telephone interventions provide a novel method to provide mental health support.

“Telephone interventions provide a novel method to provide mental health support”

• cites Kaplan, 2006Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Psychological therapies delivered by telephone were as effective as face-to-face treatment.

“Several studies have demonstrated that psychological therapies delivered by telephone were as effective as face-to-face treatment (Mohr et al., 2008)”

• cites Mohr et al., 2008Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Blinding research personnel to treatment allocation reduces selection bias by preventing researchers from influencing group assignment.

“Blinding research personnel to treatment allocation reduces selection bias by preventing researchers from influencing group assignment, whether consciously or unconsciously (Akobeng, 2005)”

• cites Akobeng, 2005Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Selection bias can compromise study design and reduce reliability of the results.

“Selection bias can compromise study design and reduce reliability of the results (Akobeng, 2005).”

• cites Akobeng, 2005Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Inadequate blinding of participants can affect expectations, reporting of symptoms, and increase trepidation and study withdrawal.

“Inadequate blinding of participants can affect expectations, reporting of symptoms, and can increase their trepidation and study withdrawal (Devereaux et al., 2002;Schulz & Grimes, 2002).”

• cites Devereaux et al., 2002Not checked yet

  Not checked in the free run.

• cites Schulz & Grimes, 2002Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Using motivation techniques like MI may be a novel approach in engaging patients to attend further treatment.

“Using motivation techniques like MI may be a novel approach in engaging patients to attend further treatment.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Most included studies stated that the MI interviews were assessed.

“despite most studies stating that the MI interviews were assessed.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Screening and referral to treatment in settings other than those concerned with mental health may be viable.

“Screening and referral to treatment in settings other than those concerned with mental health may be viable.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

The studies included in this review were generally sound in design and execution.

“The studies included in this review were generally sound in design and execution.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

MI can be used by clinicians in all health settings to promote and facilitate behaviour change, particularly for patients resistant or ambivalent to change.

“Although the review focused on mental health settings, MI can be used by clinicians in all health settings to promote and facilitate behaviour change particularly for patients that are resistant or ambivalent to change.”

• AssumptionRests on an assumption: Each interaction with a patient, no matter how brief, represents an opportunity for health promotion.

Probably not a substantial claim: may not pass S3 (load-bearing).

Where allocation concealment was inadequate or unclear, studies reported larger treatment effects, with odds ratios increased by 30% to 41%.

“Literature reviews have also indicated that where allocation concealment was either inadequate or unclear, studies reported larger treatment effects where OR can be increased by 30% to 41% (Schulz et al., 1995).”

• cites Schulz et al., 1995Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Measuring process ensures MI is delivered according to treatment principles and can measure variability between therapists.

“Measuring process ensures the MI is delivered according to the treatment principles and can measure variability between therapists (Rubin et al., 2001).”

• cites Rubin et al., 2001Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Sample size has an important influence on study quality.

“Sample size has an important influence on study quality (Smith & Beh, 2012)”

• cites Smith & Beh, 2012Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Type 2 error could occur where researchers report nonsignificant results of an intervention which may be effective.

“Similarly, type 2 error could occur where researchers report nonsignificant results of an intervention which may be effective (Borrelli, 2011).”

• cites Borrelli, 2011Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Telephone-delivered MI is a viable low-cost option in promoting continued mental health care and is a novel approach to mental health treatments.

“Telephone-delivered MI is a viable low-cost option in promoting continued mental health care and is a novel approach to mental health treatments.”

May combine two claims.

Trained therapists can diminish an individual's ambivalence regarding health-promoting behaviours and help motivate change due to MI principles.

“Due to MI principles, trained therapists can diminish an individual's ambivalence regarding health-promoting behaviours and help motivate change.”

May combine two claims.

Small sample sizes have limited power to detect the true effect of an intervention, which may lead to false positive results.

“with small sample sizes having limited power to detect the true effect of an intervention, which may lead to false positive results (Button et al., 2013).”

• cites Button et al., 2013Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Process measures are underreported in the research and difficult to ascertain causal links to outcomes.

“As process measures are underreported in the research and difficult to ascertain the causal links to outcomes”

Probably not a substantial claim: may not pass S2 (checkable).

Outcome measures like treatment attendance can indicate the quality and success of the intervention.

“outcome measures like treatment attendance can indicate the quality and the success of the intervention.”